[Boelie_Elzen,_Frank_W._Geels,_Ken_Green_(Editors)(BookFi.org).pdf

System Innovation and the Transition toSustainability

System Innovationand the Transition toSustainabilityTheory, Evidence and Policy

Edited by

Boelie Elzen

Senior Researcher, School of Business, Public Administrationand Technology, University of Twente, The Netherlands

Frank W. Geels

Assistant Professor, Department of Technology Management,Eindhoven University of Technology, The Netherlands

Ken Green

Professor of Environmental Innovation Management,Manchester Business School, University of Manchester, UK

Edward ElgarCheltenham, UK Northampton, MA, USA

Boelie Elzen, Frank W. Geels, Ken Green 2004

All rights reserved. No part of this publication may be reproduced, stored ina retrieval system or transmitted in any form or by any means, electronic,mechanical or photocopying, recording, or otherwise without the priorpermission of the publisher.

Published byEdward Elgar Publishing LimitedGlensanda HouseMontpellier ParadeCheltenhamGlos GL50 1UAUK

Edward Elgar Publishing, Inc.136 West StreetSuite 202NorthamptonMassachusetts 01060USA

A catalogue record for this bookis available from the British Library

Library of Congress Cataloguing in Publication Data

System innovation and the transition to sustainability : theory, evidence andpolicy / edited by Boelie Elzen, Frank W. Geels, Ken Green.

p. cm.Includes index.1. Sustainable development. 2. Environmental policy. 3. System theory.

I. Elzen, Boelie, 1953 II. Geels, Frank W., 1971 III. Green, Kenneth.

HC79.E5S967 2004338.927dc22

ISBN 1 84376 683 3 2004047069

Printed and bound in Great Britain by MPG Books Ltd, Bodmin, Cornwall

Contents

List of figures viiList of tables viiiList of contributors ixList of abbreviations xvPreface xviiiForeword xx

1 General introduction: system innovation and transitions tosustainability 1Frank W. Geels, Boelie Elzen, Ken Green

PART I THEORETICAL EXPLORATIONS OF TRANSITIONS

2 Understanding system innovations: a critical literature reviewand a conceptual synthesis 19Frank W. Geels

3 Socio-technological regimes and transition contexts 48Frans Berkhout, Adrian Smith and Andy Stirling

4 Sustainability, system innovation and the laundry 76Elizabeth Shove

PART II EMPIRICAL EXAMPLES OF TRANSITIONS

5 A transition towards sustainability in the Swiss agri-food chain(19702000): using and improving the multi-level perspective 97Frank-Martin Belz

6 The transition from coal to gas: radical change of the Dutchgas system 114Aad Correlj and Geert Verbong

PART III TRANSITION POLICY

7 Managing the transition to sustainable mobility 137Ren Kemp and Jan Rotmans

v

8 Getting through the twilight zone: managing transitionsthrough process-based, horizontal and interactive governance 168Geert R. Teisman and Jurian Edelenbos

9 Bounded socio-technical experiments (BSTEs): higher orderlearning for transitions towards sustainable mobility 191Halina Szejnwald Brown, Philip J. Vergragt, Ken Green,Luca Berchicci

PART IV TOOLS FOR TRANSITION POLICY ANDEMPIRICAL ILLUSTRATIONS

10 Managing experiments for transition: examples of societalembedding in energy and health care sectors 223Sirkku Kivisaari, Raimo Lovio, Erja Vyrynen

11 Socio-technical scenarios as a tool for transition policy: anexample from the traffic and transport domain 251Boelie Elzen, Frank W. Geels, Peter S. Hofman and Ken Green

12 Conclusion. Transitions to sustainability: lessons learned andremaining challenges 282Boelie Elzen, Frank W. Geels and Ken Green

Index 301

vi Contents

Figures

1.1 System optimization versus system innovation 21.2 Typology of innovations 52.1 Socio-technical system for modern car-based transportation 202.2 The multi-actor network involved in socio-technical systems 342.3 Alignment of trajectories in socio-technical regimes 352.4 Multiple levels as a nested hierarchy 362.5 A dynamic multi-level perspective on system innovations 382.6 Integration of different literatures in the multi-level perspective 403.1 Four transition contexts and transformation processes 674.1 Laundering as a system of systems 824.2 Horizontal and vertical dimensions of socio-technical change 854.3 Modes of integration 875.1 A transition in the Swiss agri-food chain, 19702000 986.1 The evolution of Dutch energy use 1156.2 The evolution of energy use in domestic households 1267.1 Four phases of transition 1407.2 Current policy versus transition management 1457.3 The cycle of transition management 1467.4 Transition as a goal-seeking process with multiple transition

images and goals 1488.1 Three democratic systems 1849.1 Mitka, May 2002 2029.2 Mitka two-wheeled concept 21010.1 The key actors in societal embedding 22610.2 Societal embedding as a learning loop 22710.3 The principle of the ESCO concept 23010.4 Process description: societal embedding of ESCO concept in

municipalities 23210.5 The development of ESCO operations in 19942003 23610.6 Driving forces for diffusion of ESCO practices in

municipalities 23710.7 Driving forces for diffusion of diabetes self-management

systems 24510.8 Two basic modes of societal embedding 24811.1 Different transition policies in different phases 256

vii

Tables

1.1 Different policy paradigms 9

8.1 Transitions seen as revolutions and evolutions 1759.1 Illustrations of higher-order learning 1989.2 Actors dilemmas: individual versus organization 2049.3 Actors dilemmas: broad support versus congruency of goals 2059.4 Unresolved barriers to Mitka adoption 2079.5 Results of surveys in Mitka development 208

viii

Contributors

Frank-Martin Belz is Professor of Brewery and Food IndustryManagement at the Technical University of Munich. He is located at theCentre of Life and Food Sciences in Weihenstephan. Frank-Martin Belzstudied Business Administration at the universities of Giessen andMannheim (Germany), majoring in marketing/consumption. In 1995 hereceived his PhD at the University of St Gallen (Switzerland). His maininterests of research are sustainability marketing and consumption. Overthe past decade he focused these efforts on the domains of food/eating,houses/living and mobility/moving, participating in various interdiscipli-nary and international research projects.

Luca Berchicci is a PhD researcher at Delft University of Technology in thedepartment of the Design for Sustainability programme. His researchfocuses on environmental product development within networks and entre-preneurial firms. He graduated at Urbino University in Geology and aftera period working in geophysical and geological matters he decided to goback to study joining the EAEME postgraduate course in EnvironmentalManagement at Polytechnic University of Athens and at ErasmusUniversity of Rotterdam. On that occasion during the Ecodesign lectureshe had the opportunity to meet Professor Brezet and J.C. Diehl, and shortlyafterwards became a research fellow in their department.

Frans Berkhout is director of the ESRC Sustainable TechnologiesProgramme, and Senior Fellow and Head of the Environment and EnergyProgramme at SPRU Science and Technology Policy Research at theUniversity of Sussex. He has extensive research and research managementexperience across a number of fields. His early research was concerned withthe economic, political and security aspects of the nuclear fuel cycle andradioactive waste management, with special emphasis on the internationalcontrol of fissile materials (plutonium and highly enriched uranium). Morerecent work has been concerned with technology, policy and sustainability,with special emphasis on the links between technological innovation andenvironmental performance in firms, the measurement of sustainabilityperformance, futures scenarios studies, business adaptation to environ-mental change, and policy frameworks for innovation and the environment.

ix

Halina Szejnwald Brown is Professor of Environmental Health and Policyat Clark University, Worcester, MA, USA. She received a PhD degree inchemistry from New York University. Prior to joining Clark University,Brown was a chief toxicologist for the Massachusetts Department ofEnvironmental Protection. Browns research focuses on environmentalregulatory regimes in the USA and Europe, the use of science and infor-mation in public policy, and social learning through technological innova-tion for sustainability. Brown has authored over 40 articles and two booksand served on numerous state and national advisory panels, including theNational Academy of Science, Environmental Protection Agency, theMassachusetts Toxic Use Reduction Institute, National ScienceFoundation, and the American Association for the Advancement ofScience. She is Fellow of the Society for Risk Analysis, and Fellow of theAmerican Association for the Advancement of Science.

Aad Correlj is Associate Professor at the Section Economics ofInfrastructures of the Faculty of Technology, Policy and Management ofthe Delft University of Technology. He is a research fellow with theClingendael International Energy Programme (CIEP). After studyingPolitical Science (International Relations and European Law) at theUniversity of Amsterdam, he wrote a PhD thesis at the Centre forInternational Energy Studies (EURICES) at the Erasmus University inRotterdam. Correlj has been involved in academic research, teaching andadvising on many energy-related issues, including oil and gas markets, eco-nomic restructuring and (sustainable) energy policy.

Jurian Edelenbos is assistant professor at the Centre for PublicManagement, Erasmus University, Rotterdam, the Netherlands. He haspublished on interactive governance, process management, publicprivatepartnership, transition management, and trust in complex interorganiza-tional cooperation. He does research mainly in the domain of urban plan-ning and infrastructure.

Boelie Elzen is senior researcher at the School of Business, PublicAdministration and Technology, University of Twente, Enschede, theNetherlands. His general research interest is in understanding the dynam-ics of socio-technical change and using these insights to develop sugges-tions on how to tackle the societal problems related to these changeprocesses. Over the past decade he has focused these efforts on the domainof passenger mobility, worked on various relevant EU research projectsand acted as consultant to various institutions in the field, in theNetherlands as well as abroad.

x Contributors

Frank W. Geels is Assistant Professor at the Department of TechnologyManagement, Eindhoven University of Technology, the Netherlands. Hisgeneral research interest is in understanding the dynamics of technologicaltransitions and system innovations. To understand these dynamics he hasdeveloped a conceptual perspective, using insights from innovation studiesand the sociology of technology. He also does case-studies from differentdomains (for example, transport, manufacturing, hygiene), mainly histori-cal, to test and refine the perspective. He has also done prospective work onsocio-technical scenarios, using the same perspective.

Ken Green is Professor of Environmental Innovation Management in theManchester Business School at the University of Manchester. He teachesand researches in technology and innovation management, with a stronginterest in environmental issues. He is director of the schools Centre forResearch on Organisations, Management and Technical Change(CROMTEC). He is also a director of the newly-formed Institute ofInnovation Research and is associated with the Tyndall Centre for ClimateChange Research. Professor Greens current research interests are in thesocio-economic analysis of technological development, especially withregard to environmental influences on innovation. Current research con-tracts are with the Tyndall Centre, on the influence of long-term techno-logical change on greenhouse gas emissions, and the ESRC, onsustainability in food systems.

Peter S. Hofman is a senior research associate of the Center for CleanTechnology and Environmental Policy at the University of Twente inEnschede, the Netherlands. His research centres on the role of policy andinnovation in transition paths towards more sustainable production andconsumption. He currently holds a post-doctoral position, funded byNWO (the Dutch national science foundation), in which he is involved inanalysis of socio-technical change related to the energy system and thedevelopment of socio-technical scenarios.

Ren Kemp is senior research fellow at Maastricht Economic ResearchInstitute on Innovation and Technology (MERIT) from MaastrichtUniversity and senior advisor of TNO-STB. He is an expert on thetopic of innovation for the environment, about which he has publishedseveral books and articles (see www.meritbbs.unimaas.nl/rkemp). Hiscurrent work is on transitions to sustainability, focusing on how thesetransitions may be managed. He has been consultant for the EuropeanCommission, OECD and Dutch government and can be contacted [email protected].

Contributors xi

Sirkku Kivisaari is senior researcher at VTT Technology Studies, VTTTechnical Research Centre of Finland. Her research relates to future-ori-ented technology assessment and especially to developing the approach ofsocietal embedding for promoting dialogue between producers, users andsocietal actors in shaping new health services and technologies. Her currentwork focuses on facilitating publicprivate partnerships for developing andintroducing innovations provoked by societal concerns for wellbeing of theageing society and for a cleaner environment.

Raimo Lovio is Professor of Innovation and EnvironmentalManagement at the Department of Management, the Helsinki Schoolof Economics, Finland. In recent years he has studied energy sectordevelopment and innovations from the point of view of climatechange. In addition, his current work focuses on the globalization oflarge Finnish companies in terms of innovation activities and corporateresponsibility, and the role of environmental management systems andreporting in improving the environmental performance of businessorganizations.

Jan Rotmans is one of the founders of Integrated Assessment (IA), and hasoutstanding experience in IA modeling, scenario-building, uncertaintymanagement and transition management. During the past 15 years he hasled a diversity of innovative projects in the field of climate change, globalchange, sustainable development and transitions. Since 1992 Jan Rotmanshas had a professorship at Maastricht University, and since 1998 a full pro-fessorship on Integrated Assessment. He is founder and director of theInternational Centre for Integrative Studies (ICIS) (1998) at MaastrichtUniversity. He is Vice-Chairman of the European Forum on IntegratedEnvironmental Assessment (EFIEA), Vice-President of the IntegratedAssessment Society (TIAS), and founder of the Dutch KnowledgeNetwork on System Innovations: transitions towards a sustainable society(KSI).

Elizabeth Shove is a reader in the Department of Sociology at LancasterUniversity. She has worked on different aspects of environmental sociol-ogy, from energy and buildings to the domestic freezer. Her recent book,Comfort, Cleanliness and Convenience: the Social Organization ofNormality, takes these ideas forward with reference to key areas of every-day life and ordinary consumption. Elizabeth is currently working on pro-jects relating to the kitchen, the bathroom and the future of comfort, thesebeing important arenas in which to explore images, practices and tech-nologies, and the dynamic relation between them.

xii Contributors

Adrian Smith is a social scientist who analyses relationships between tech-nology, society and sustainable development. He has done research for avariety of public and non-governmental organizations, including theEconomic and Social Research Council, the European Commission,the Environment Agency, the Department of the Environment, theDepartment of Trade and Industry, the Sustainable DevelopmentCommission and the Institute for Prospective Technological Studies. He isa member of the editorial board for two journals: Sustainable Development;and Theomai Revista de Sociedad, Naturaleza y Desarrollo. He also sitson the advisory board of the International Bibliography of the SocialSciences.

Andy Stirling is a senior lecturer at SPRU science and technology policyresearch, at the University of Sussex. He has a background in astrophysics,a Masters in social anthropology and archaeology (Edinburgh) and a DPhilin technology policy (Sussex). Formerly working for GreenpeaceInternational, he later served on their Board. His research focuses on tech-nological risk and technology choice in areas spanning the nuclear, energy,chemicals, medical and biotechnology sectors. In particular, he is activelyinvolved in developing more participatory and precautionary approachesin technology policy and in the analysis of strategic issues like flexibility,resilience and diversity. He has served on a number of UK and EU policyadvisory committees, including the EUs Energy Policy ConsultativeCommittee, the UK Advisory Committee on Toxic Substances and the UKGM Science Review Panel.

Geert R. Teisman is Professor in Decision Making and Complexity Theoryand Chair of the Centre for Public Management at the Erasmus Universityin Rotterdam. His research topics are intergovermental cooperation,publicprivate partnership and management of joint decision making,mainly in the areas of transport, spatial and economic developments andenvironmental affairs.

Erja Vyrynen is researcher at VTT Technology Studies, VTT TechnicalResearch Centre of Finland. Her areas of expertise are technology fore-sight and societal embedding of innovations. Her earlier activities haveranged from urban planning to environmental impact assessment. She haslately been involved in developing Finnish technology foresight practiceson the basis of European experience. In her recent study on ageing of thepopulation as a challenge for technology foresight and innovations, she alsodiscusses the role of national innovation policy in developing social inno-vations.

Contributors xiii

Geert Verbong is Associate Professor in Technology and SustainabilityStudies, Department of Technology Management, Eindhoven Universityof Technology (EUT). His main area of research is on system innovations,in particular energy systems, and on implementation strategies for renew-able energy technologies. He has been an editor and researcher in largeprojects on the history of technology in the Netherlands in the nineteenthand twentieth centuries. Currently his research focuses on electricitysystems in Europe and on biomass. He teaches Technology Assessment,System Innovations and Strategic Niche Management in the STS pro-gramme of EUT, the MSc programme Sustainable EnergyTechnologies and other EUT engineering programmes.

Professor Philip J. Vergragt is a visiting senior fellow at Tellus Institute,presently on leave from Delft University of Technology in the Netherlandswhere he holds a professorship. His research focuses on societal transitionstowards sustainable mobility and sustainable consumption, with specialinterest in the questions of infrastructure, culture, stakeholder perspectivesand participation, and social learning. He facilitates visioning and back-casting workshops, by a methodology developed in the STD (SustainableTechnological Development) Programme, and the SusHouse project.

xiv Contributors

Abbreviations

ANT Actor-Network TheoryANWB (Dutch) motoring associationAT Appropriate/Alternative TechnologyAVG Automatic Vehicle GuidanceBCM billion cubic metresBOM Brabant Development CorporationBPM Bataafse Petroleum Maatschappij (subsidiary of Shell in

the Netherlands)BSTE Bounded Sociotechnical ExperimentsCEV city electric vehiclesCIEP Clingendael International Energy ProgrammeCMS City Mobility SystemCROMTEC Centre for Research on Organizations Management and

Technical Change, Manchester Business School, (UK)DSM Dutch State MinesDFS Design for SustainabilityDTO Sustainable Technology Development programme

(Dutch)EAEME postgraduate course in Environmental Management at

the Polytechnic University of Athens and ErasmusUniversity of Amsterdam

ECMT European Conference of Ministers of TransportEFIEA European Forum on Integrated Environmental

AssessmentESCO Energy Service CompanyEURICES Centre for International Energy Studies, Erasmus

University, RotterdamEUT Eindhoven University of TechnologyFCB Fuel cell busGPS Global Positioning SystemHEV hybrid electric vehicleIA Integrated AssessmentICIS International Centre for Integrative Studies, Maastricht

University, NetherlandsICT Information and Communication Technology

xv

IHDP International Human Dimensions ProgrammeIRA increasing returns to adoptionKNV freight transport organization in the NetherlandsKSI (Dutch) Knowledge Network on System InnovationsLDV long-distance vehicleLTS Large Technical SystemsMC Mobility CentreMERIT Maastricht Economic Research Institute on Innovation

and Technology, Maastricht University, NetherlandsMTI Ministry for Trade and Industry (Finland)NAM Nederlandse Aardolie Maatschappij, partnership of

BPM (Shell) and Standard Oil Company of New JerseyNEPP National Environmental Policy Plan (Netherlands)NS National Rail Company, NetherlandsNVVP National Plan for Traffic and Transport, NetherlandsNWO Dutch Research Councilpkt passenger kilometres travelledRAI, BOVAG represent car dealers and garages in the NetherlandsSCOT Social Construction of TechnologySGB State Gas Company (Netherlands)SMEC Social Management of Environmental ChangeSNAM Societ Nazionale Metanodotti (National Methane

Company), ItalySNM Strategic Niche ManagementSPRU Science and Technology Policy Research, University of

SussexSROG Commission Cooperation Regional Organizations Gas

SupplySTF Sustainable Texel FoundationSTSc Socio-technical ScenarioSTS Science, Technology and SocietyTBM Faculty of Technology Policy and Management,

University of DelftTEMO Texel Own Mobility SystemTEP Technoeconomic paradigm3VO traffic safety organization in the NetherlandsTIAS The Integrated Assessment Society (Netherlands)TNO Organization for applied research in technological

innovation in industryV&W Ministry of Transport, Public Works and Water

Management (the Netherlands)VEGIN United gas companies in the Netherlands

xvi Abbreviations

VROM Ministry of Housing, Spatial Planning and theEnvironment, Netherlands

VSN regional transport providers in the NetherlandsVTT Technical Research Centre of Finland

Abbreviations xvii

Preface

Since the publication of the Brundtland report in 1987, the goal of sus-tainability has increasingly gained the attention of a variety of societalactors, including public authorities, NGOs, consumer groups and indus-trial firms as well as researchers in a wide range of disciplines. At thegeneral level, there is widespread consensus that various characteristics ofmodern societies are not sustainable and should change. When things getmore prescriptive, however, many feel that the goals of sustainability seemto clash with other vital societal interests.

In recent decades, impressive results have been achieved in the environ-mental aspect of sustainability, for example by curbing the emissions of avariety of pollutants. Nonetheless, many feel that achieving the broadergoals of sustainability is still remote since many problems appear extremelydifficult to tackle, such as obtaining large reductions in the emission ofgreenhouse gases. Furthermore, the scope of the term of sustainability hasbecome broadened to include a variety of goals, including a healthy envi-ronment, a healthy society and a healthy economy. To achieve this multi-tude of targets we seem to need fundamental changes, and these changesare denoted by terms like system innovation, transition and industrialtransformation.

Across the world, researchers from different disciplinary backgroundshave begun to try to understand the processes underlying these changes andpolicy makers have begun to use these insights. In the Netherlands, forexample, various ministries have set up so-called transition teams whowrestle with the issue of how to set in motion fundamental changes towardsachieving sustainability. This suggests a need to exchange insights, experi-ences and views between a divergent research community and policymakers.

This led to a Dutch initiative to organize an international workshop onTransitions Towards Sustainability Through System Innovation, held atthe University of Twente in the summer of 2002. The workshop was fundedby the RMNO (The Dutch Advisory Council for Research on Nature andEnvironment), the Dutch National Council for Agricultural Research(Innovatienetwerk Groene Ruimte en Agrocluster), the Dutch Ministry ofHousing, Spatial Planning and the Environment, the IndustrialTransformation Project of the International Human Dimensions

xviii

Programme (IHDP IT), the Greening of Industry Network, the DutchNational Initiative for Sustainable Development (NIDO) and theUniversity of Twente.

The workshop was organized by an international steering committee andselected participants came from ten different countries. They includedresearchers with various disciplinary backgrounds as well as policy makers.The main goal of the workshop was to seek some common ground amongstthe heterogeneity of approaches, and define an agenda for further work.This book contains a selection of ten papers that were prepared for theworkshop and fuelled the discussions; it includes a general introductionand a conclusion that teases out some general findings.

We would very much like to thank our sponsors for making it possible toorganize this workshop and all participants for their contributions as eitherauthors or commentators and for their participation in the discussionsmaking the workshop a success.

Boelie ElzenFrank W. GeelsKen Green

WORKSHOP PARTICIPANTS

Colette Alma, Peter Aubert, Theo Beckers, Frank-Martin Belz, FransBerkhout, Halina Brown, Tine Bruland, Joske Bunders, Maurits Butter,Aad Correlj, Frans Duijnhouwer, Boelie Elzen, Gertjan Fonk, FrankGeels, Ken Green, John Grin, Rob Hoppe, Jorge Islas, Klaus Jacob, UlrikJrgensen, Ren Kemp, Sirkku Kivisaari, Derk Loorbach, Rob Maas, ArieRip, Harald Rohracher, Jan Rotmans, Johan Schot, Elizabeth Shove, RuudSmits, Geert Teisman, Andrew Tylecote, Pier Vellinga, Geert Verbong,Frans Vollenbroek, Matthias Weber, Anna Wieczorek, Jan de Wilt.

Preface xix

Foreword

On behalf of the sponsors of the international workshop on TransitionsTowards Sustainability Through System Innovation I am happy to recom-mend to you the edited volume of the most interesting research results andideas presented at this workshop.

Global environmental change poses an unprecedented internationalchallenge for 21st century societies since it requires a radical change in theway human needs in the field of energy, food, water and mobility are met.It calls for a transformation of our current consumption and productionpatterns as well as a transformation of incentive structures and the institu-tions that shape the relationship between the two. Such a proactiveapproach is based on the understandings of system analysis, system beingdefined as a set of inter-related economic activities and actors and flows ofgoods and services. For system change to be effective, it needs attention inall aspects of life: technology, institutions, economy, and the socio-culturalsphere. Because of this complexity, it is not surprising that a change to amore sustainable system will require a long time at least one generation.

For that reason, research into societal transformations that have thepotential to decouple economic development from environmental burdenhas become the focus of many research institutes worldwide. Two types ofresearch activities can be distinguished: one is focused on understandingthe dynamics of past transitions that, very often, occurred without delib-erate planning; the second type of research focuses on the possibilities ofsteering societal changes towards sustainability.

This book is one of the first to present the state of the art in knowledge oftransitions towards sustainability through system innovation. Even thoughknowledge in this field is still in its infancy, we are starting to recognize the foun-dation of a new field of research with a new set of definitions and approaches.I very much hope this book will serve as a good starting point for those whowant to further expand this field. The book is primarily meant for those whoare curious about how transformations take place and what problems societiesface when they want to steer these great changes in a desired direction.

Pier VellingaChair of the Scientific Steering Committee of the Industrial Transformation Projectof the International Human Dimensions Programme on Global EnvironmentalChange (IHDP IT)

xx

1. General introduction: systeminnovation and transitions tosustainabilityFrank W. Geels, Boelie Elzen, Ken Green

Modern societies face structural problems in several sectors. In the energysector there are problems related to oil dependency, reliability, and CO2 andNOx emissions. The transport system suffers from congestion, air pollution(particulates, NOx), energy use and CO2 emissions. Cattle farming suffersfrom manure disposal problems, ammonia emissions and diseases like BSEand foot and mouth disease. These problems are deeply rooted in socialproduction and consumption patterns.

Since the 1980s, much effort has been made to solve problems withproduct and process innovations. Cleaner products and processes have beendeveloped alongside the application of end-of-pipe solutions. Sometimesthese innovations have led to substantial improvements in environmentalefficiency, such as in the case of automobile catalysts which greatly reducedtailpipe-emissions of pollutants. The focus in these cases has been onchanging some technological artefact.

Substantial improvements in environmental efficiency (a Factor 2 is ageneral average) may still be possible with innovations of an incrementalkind. But larger jumps in environmental efficiency (possibly by a Factor 10)may only be possible with system innovations. The promise of transitionsto sustainability via system innovations is schematically represented inFigure 1.1. Such transitions to sustainability require changes from, forexample, one transport system to another or from one energy system toanother. Such system innovations not only involve new technological arte-facts, but also new markets, user practices, regulations, infrastructures andcultural meanings.

Because of its sustainability potential there is increasing interest frompolicy makers, NGOs and large firms in transitions and system innova-tions. The Stockholm Environment Institute, for instance, has publisheda book on the Great Transition (Raskin et al., 2002). The AmericanNational Research Council (1999) and the Dutch Research Council

1

(NWO) have made the study of transitions part of their research port-folio. The IHDP research programme (International Human DimensionsProgramme on Global Environmental Change) has a Project on IndustrialTransformation (similar in meaning to transitions). The Dutch govern-ment gave transitions a central place in its fourth National EnvironmentalPolicy Plan (VROM, 2001) and has established transition teams withinvarious ministries.

To link with and to feed this growing interest, this book explores howsystem innovations come about and how policymakers might influencethem.

DELINEATING THE TOPIC OF ANALYSIS:TRANSITIONS AND SYSTEM INNOVATIONS

In Websters Dictionary the term transition is defined as a passage fromone state, stage, subject, or place to another or a movement, development,or evolution from one form, stage, or style to another. The states/formshave certain internal characteristics, which give them coherence and stabil-ity. The notion of a transition also has the connotation of rapid change, ajump from one state to another.

Transitions can occur on different levels, depending on the unit of analy-sis. An example at the level of society as a whole is the transition from

2 Introduction

5 10 20 Time horizon (years)

Factor 10

Factor 5

Factor 2

Impr

ovem

ent i

n en

viro

nmen

tal e

ffic

ienc

y

Function innovation= new system

Partial systemredesign

System optimimization

Source: Weterings et al., 1997.

Figure 1.1. System optimization versus system innovation

hunter-gatherer society to urban society. Another example is the transitionfrom rural to industrial society. At a lower level, there are transitions in soci-etal functions such as transport, communication, housing, feeding, energysupply and use, and recreation. Examples are the transition in transportsystems from horse-and-carriage to automobile, or the transition from tele-graph to telephone. There are also transitions at the level of organizationsand firms, for example the transition from punched card machines to com-puters within IBM (Chandler, 2001) or the transition of DSM (Dutch StateMines) from coal mining via bulk chemicals to fine chemicals. This bookfocuses on a specific type of transition, notably transition at the level ofsocietal functions.

What is it that changes at this level during transitions? In the way we usethe term, these transitions involve changes in socio-technical systems. Thesecomprise a cluster of elements, including technology, regulations, userpractices and markets, cultural meanings, infrastructure, maintenance net-works and supply networks. Technology plays an important role in fulfill-ing societal functions, but its functioning depends upon its relationship tothe other elements. Technologies realize functionalities in concrete usercontexts, which are made up of users, their competencies, preferences, cul-tural values and interpretations. User contexts are also shaped by a varietyof existing artefacts and infrastructures (for example, road infrastructures,electricity networks), and regulations.

Technologies also need to be produced, distributed and tuned withexisting user contexts. This requires aspects such as technological knowl-edge, machines, skilled labour, capital, natural resources and components,and distribution networks. Although these supply and demand aspects canbe distinguished analytically, they are mutually dependent in practice. Tohighlight this interrelatedness, we use the term socio-technical systems.Transitions at the societal level then involve a change from one socio-tech-nical system to another, that is a system innovation.

This way of delineating the unit of analysis has several implications.Firstly, it means that the focus is wider than just an industry or a sectoralsystem of innovation. There has been some attention paid in the past tothe emergence of new industries (Van de Ven and Garud, 1989), howindustries coevolve with government and universities in so-called triplehelix dynamics (Etzkowitz and Leydesdorff, 2000), and how firms, publicauthorities and universities work together in innovation systems or inno-vation communities (Breschi and Malerba, 1997; Malerba, 2002; Lynn etal., 1996). These approaches, however, mainly look at the supply-side andthe production of innovations. They take the user side for granted ornarrow it down to the market which functions as a neutral selection envi-ronment.

Introduction 3

There is another body of literature which shows that users do more thanjust buy and adopt (new) technologies. Cultural studies and social studiesof technology have found that users have to domesticate new technol-ogies to fit existing user contexts. This involves symbolic and practicalwork, in which users integrate the artefact into their user practices, andcognitive work, which includes learning about the artefact (Lie andSrensen, 1996; Du Gay et al., 1997). This fits an emerging trend in inno-vation studies and science and technology studies, in which more attentionis paid to the role of users in innovation and technological development(see Schwartz-Cowan, 1987; Kline and Pinch, 1996; Eggerton, 1999;Coombs et al., 2001; Oudshoorn and Pinch, 2003). These observationsimply that system innovations not only involve changes in industries, firmsand technical knowledge, but also changes in user contexts and symbolicmeanings.

This book acknowledges this and aims to bring together two bodies ofliterature which have remained relatively separate so far: on the onehand evolutionary economics, innovation studies and innovation systemapproaches and, on the other hand, cultural studies, and science and tech-nology studies.

A second implication is that system innovations appear as a particularkind of innovation. To illustrate this, we use the innovation typology ofAbernathy and Clark (1985), but widen it. They distinguish two dimen-sions. The first dimension consists of linkages between a firm and itscustomers, including channels of distribution and service, customer appli-cations and customer knowledge. The second dimension relates to thetechnological and production competences of a firm, including pro-duction systems, skills, technical knowledge and supplier relations.Combining these two dimensions results in four types of innovations (seeFigure 1.2):

1. architectural: disrupts existing technology and linkages with users;2. niche creation: conserves existing technology but breaks linkages with

users; new markets are explored with existing products;3. incremental: conserves both existing technology and users;4. revolutionary: disrupts technology but conserves user linkages (same

markets).

Abernathy and Clark developed their innovation typology primarily todetermine the consequences of different kinds of innovations for firms.While their point about linkages and alignments between elements is import-ant, their focus on firms is too limited for our purposes as disruptions occuron a much wider scale during system innovations. Hence, system innovations

4 Introduction

can be described as architectural innovations writ large, because theyinvolve substantial changes on the supply side and on the user side. The termalso highlights that system innovations are not about component changes,but about changes in the entire architecture or structure of socio-technicalsystems. Without changes on the user side, technological discontinuitiesare better described as technological revolutions, which do not changefunctionalities.

A third implication is that system innovations are multi-actor processes.This not only denotes interactions between actors within a societal group (forexample, industry, user group, scientific community, policy community), butalso interactions between societal groups. A range of societal groups orstakeholders is involved in system innovations: firms, suppliers, universitiesand knowledge institutes, public authorities, public interest groups, users.Their activities create and maintain elements of socio-technical systems. Thesocietal groups have their own perceptions of the future, values and prefer-ences, strategies, and resources (money, knowledge, contacts). Althoughthese societal groups have some degree of autonomy, they are also related toeach other and interpenetrate each other (Stankiewicz, 1992).

Their activities are to some degree aligned, and it is this that gives socio-technical systems their stability and a recognizable state or form. Withinsystems, innovations still take place but they are usually of an incrementalnature, leading to trajectories in technical development, policies, infrastruc-tures and demand. As long as these trajectories are aligned, socio-technicalsystems are stable. This stability is not the result of an overarching ration-ality or force by an all-powerful actor. Instead, stability is the emergentoutcome of many activities of many actors. Stability need not be harmo-nious. There may be tensions and conflicts of opinion about a range of

Introduction 5

Creationmarket niche

Architectural

IncrementalRadical,

revolutionaryTechnology

Conserve/entrenchexisting competences

Disrupt/obsoleteexisting competences

Markets/customerlinkages

Disruptexisting/create

new linkages

Conserve/entrench existing

linkages

Source: Abernathy and Clark, 1985.

Figure 1.2. Typology of innovations

matters, such as which problems should be highest on the problem agenda,which directions are most promising for solving a problem, or how resourcesshould be allocated. When these tensions become pressing, a system maylose its stability, creating opportunities for change. But usually tensionsremain manageable.

This discussion implies that, for our purposes, transitions or systeminnovations have the following main characteristics. First, they developin a coevolutionary way. They involve changes in both the supply side (intechnology, knowledge, industry structures) and the demand side (userpreferences, cultural meaning, infrastructure). Second, they are architec-tural innovations writ large, involving changes in the elements and struc-ture of socio-technical systems. Third, they are multi-actor processes,involving a wide range of societal groups. A fourth characteristic thatfollows from a wide range of historical studies is that they unfold withina long timescale, possibly of the order of several decades (see Geels,2002).

ACADEMIC RELEVANCE

Given the specifics of our interest in system innovation as described above,we identify three gaps in existing bodies of literature. The first gap concernsthe systems of innovation approach, which has emerged in the last decade.This approach investigates at different analytical levels how innovationsemerge from the coevolution of a range of elements. The levels are nationalsystems of innovation, regional systems of innovation or sectoral inno-vation systems. The main focus in the system of innovation approach is onthe functioning of systems rather than the change of systems. For instance,at the national level these studies lead to a static or comparative analysis ofthe innovative performance of different countries. In a recent overview ofsectoral systems of innovation, it was noted that one of the key questionsthat need to be explored in-depth is: how do new sectoral systems emerge,and what is the link with the previous sectoral system? (Malerba, 2002;262). This means that the topic of system innovation as we use it is under-addressed in this literature.

A second gap stems from the literature on path dependence andlock-in. In evolutionary economics, David (1985) and Arthur (1988) haveshown that path dependence plays an important role in the case of twocompeting technologies. Once one of the technologies has gained a lead,it benefits from increasing returns to adoption and creates a dominantpath. Several mechanisms cause increasing returns, such as economiesof scale, leading to lower cost, learning-by-using, network externalities,

6 Introduction

informational increasing returns, and technological interrelatedness.Because of these increasing returns a certain technology becomesentrenched while there is no guarantee it is the best one from a broadersocietal perspective. Other economists have widened the analysis, addinginstitutional aspects and user routines to the lock-in analysis (Cowan,1990; Cowan and Gunby, 1996).

Research from other disciplines has added more reasons why existingsystems are characterized by stability, inertia and lock-in. Establishedsystems may be stabilized by legally binding contracts (Walker, 2000).Actors and organizations are embedded in interdependent networks(with suppliers or users), which represent a kind of organizational capital,and create stability through mutual role expectations. Cognitive routinesmake engineers and designers look in particular directions and not inothers (Nelson and Winter, 1982; Dosi, 1982). This can make themblind to developments outside their focus. Core capabilities can thusturn into core rigidities (Leonard-Barton, 1995). Firms have sunkinvestments and built-up capital, which they do not want to write-off(examples are investments in machines and production tools, skills andknowledge). It is difficult for established firms to switch to competence-destroying breakthroughs (Tushman and Anderson, 1986; Christensen,1997).

Existing systems are also stable because they are embedded in society.People adapt their lifestyles to them, favourable institutional arrangementsand formal regulations are created, and accompanying infrastructures areset up. The alignment between these heterogeneous elements leads totechnological momentum (Hughes, 1994). The importance of these align-ments between heterogeneous elements is highlighted in such concepts asthe techno-institutional complex (Unruh, 2000) and techno-economicnetworks (Callon, 1991). All these approaches highlight aspects of the sta-bility of existing systems but none of them addresses the issue of change andtransition from one system to another. Given all these explanations ofstability, it is a mystery how and why transitions occur. Path-dependencyliteratures may help us understand lock-in, but how can we understandlock-out?

A third gap relates to recent academic sustainability debates. There hasbeen a widening in recent years of the analytical focus, from clean productsto sustainable systems (Schot et al., 1994; Vellinga and Herb, 1999; Unruh,2000; Jacobsson and Johnson, 2000; Berkhout, 2002). In transport, energyand other systems there are promising new technologies with better envi-ronmental performance. But many of these new technologies are not (yet)taken up. This is partly for economic reasons, but there are also social, cul-tural, infrastructural and regulatory reasons. Existing systems seem to be

Introduction 7

locked in on many dimensions. Implementation of promising new envi-ronmental technologies may require other changes in user practices, regula-tion or infrastructure. Although the importance of system innovations isincreasingly emphasized in sustainability debates, there is not yet muchknown about how system innovations occur and how policy makers mayinfluence them.

GOVERNANCE

Transitions are complex, uncertain and involve multiple societal groups orstakeholders. Hence, policymakers and other decision-makers puzzle overhow to influence system innovations and how to identify possible andpromising directions for transitions. This is complicated by the awarenessthat the state is not an all-powerful and all-knowing actor in this matter.Public authorities are just one societal group among several others. Likeother groups, they have limited power, a limited cognitive perspective andlimited resources to influence system dynamics.

This observation caused a shift in policy studies from a focus on gov-ernment to governance (e.g. Kooiman, 1993; Rhodes, 1997; Kohler-Kochand Eising, 2000; Van Heffen et al., 2000). Governance means that there isdirectionality and coordination at the systems level, but that it has an emer-gent character, arising from the interaction between multiple societalgroups. Public authorities may try to influence this emergent directionality,but cannot steer it at will. This emerging governance paradigm emphasizesaspects such as policy networks, interaction between multiple societalgroups and learning processes.

This is not the only relevant policy paradigm. In policy science, threegeneral policy paradigms are distinguished (see Table 1.1): (i) the tra-ditional top-down model with a central role for (national) government andhierarchical relations, (ii) a bottom-up or market model with a large degreeof autonomy for local actors, and (iii) a governance or policy networkmodel with shared rule-making and agreements between interdependentactors with diverging values and beliefs. These three policy paradigmsdiffer not only in their basic philosophy, but also in their instruments.Formal rules and regulations are common in the command-and-controlparadigm, subsidies, taxes and (financial) incentives in the market model,and network management, learning processes, experiments and interactivepolicymaking in the third paradigm.

These policy paradigms coexist in all democratic societies with varyingdegrees of emphasis on each of them. This variety of paradigmsand instruments complicates the issue of governing transitions and system

8 Introduction

innovations. It raises questions such as these: is one policy paradigm bestsuited to influence transitions or is a mix of paradigms and instrumentsneeded? In the latter case, what should this mix look like? Is the optimalmix dependent upon specific circumstances and, if so, which ones?

Introduction 9

Table 1.1 Different policy paradigms

Classic steering Market model Policy networks paradigm (bottom up) (processes and(top-down, networks)command-and-control)

Level of Relationship is Relationship is Network ofanalysis between principal between principal actors

and agent and local actors

Perspective Centralized, Local actors Interactions hierarchical between organization actors

Characterization Hierarchical Autonomous Mutually of relationships dependent

Characterization Neutral Self Interaction of interaction implementation organization processes in processes of formulated on the basis of which information

goals autonomous and resources aredecisions exchanged

Foundation Classic political Neo-classical Sociology, innovation scientific science economy studies,disciplines neo-institutional

political science

Governance Formal rules, Financial Learning processes,instruments regulations incentives network management

and laws (subsidies, through seminarstaxes) and strategic

conferences,experiments,vision building atscenario workshops,public debates

Source: Based on De Bruijn et al., 1993: 22.

RESEARCH QUESTIONS AND LEVEL OF AMBITION

Given our dual ambition to enhance the understanding of transitionsand to stimulate the formulation of policies that guide transitionstowards sustainability there are two main research questions that drivethis book:

1. How do system innovations or transitions come about? What theoriescan be used to conceptualize (part of) their dynamics and what gapsexist in those theories? What can we learn from historical examples oftransitions?

2. How can transitions or system innovations be influenced by actors, inparticular by public authorities? What instruments and tools exist andhow should they be used?

Although this book aims high, it is not our ambition to provide the ulti-mate answers to these questions. Instead, we want to create a signpost intothis uncharted territory. System innovations are a complex topic, involvingmany kinds of actors and issues. In this book, authors from different disci-plines make their own distinctive interventions into the topic, coming at itfrom different angles and with different intellectual frameworks: innovationstudies, sociology of technology, institutional economics, history of tech-nology, policy studies, including studies of network governance, learningand the impact of regulation, innovation management and governanceapproaches, organizational studies and management of structural changeand leadership. This grouping of different disciplinary backgroundsaround a particular topic creates variety and space for interdisciplinary dis-cussion. Although different disciplines highlight different aspects of systeminnovations, they do share a common view on human actors as boundedlyrational and embedded in social networks and institutions. This meanstransitions are not and cannot be planned in advance in a rational mannerbut emerge as actors navigate their way through multiple uncertainties.These shared views provide common ground between different authors.

BOOK STRUCTURE

The book is in two sections, addressing the two main research questions.The first section is on Understanding Transitions, with one part focusingon theoretical explorations (Chapters 2, 3 and 4), and one on empiricalexamples (Chapters 5 and 6). The second section deals with InducingTransitions; it also consists of two parts, the first on transition management

10 Introduction

in general (Chapters 7 and 8) and the second on tools for transition man-agement (Chapters 9, 10 and 11).

Part I: Theoretical Explorations of Transitions

Chapter 2 by Frank Geels addresses the general question of how systeminnovations come about. Geels reviews a broad range of relevant litera-tures, concluding that they do not add up to an integrated perspectiveon system innovations. He offers a pragmatic synthesis in the form ofa multi level perspective (MLP) to analyse and explain transitionprocesses. These levels are (i) technological niches where novelties are devel-oped, (ii) socio-technical regimes and (iii) socio-technical landscape, whichcomprises a range of exogenous developments which influence regimesand niches. The main argument is that system innovations result from link-ages between processes at these multiple levels. This means that systeminnovations are not caused by a change in a single factor or driver, butare the result of the interplay of many processes and activities.

Chapter 3 by Frans Berkhout, Adrian Smith and Andy Stirling alsolooks at the general level of system innovations. They take aim at the multi-level perspective, arguing that this approach places too much emphasis onthe role of technological niches as the principal locus for regime change.Instead, they argue, there is a range of different transition contexts in whichregime change can take place. They argue that there is a greater pluralityof possible transformation pathways than suggested by the multi-level per-spective notion. They develop a four-fold typology of transition contexts,which they illustrate with brief examples.

Chapter 4 by Elizabeth Shove has a more specific focus on users andconsumption. She sees much of the current work on transitions associo-technical in its orientation as it acknowledges the institutional andpolitical processes required in support. She argues, however, that theagenda remains lopsided, skewed around provision rather than consump-tion and around the diffusion rather than the use of technological systems,tools and techniques. She seeks to recover some of that missing ground.Using the case of laundering, she argues that it is necessary to think moresystematically about the relation between consumption, provision and prac-tice. She suggests that shared understandings of normality are import-ant in this respect. Notions of what it is to be a normal and acceptablemember of society have far reaching environmental implications. Theycarry in their wake a trail of resource requirements like those associatedwith daily showering, with wearing freshly laundered clothing, withnot having a siesta, with eating imported food or with having foreignholidays.

Introduction 11

Part II: Empirical Examples of Transitions

Chapter 5 by Frank Belz presents a historical analysis of changes in theSwiss agri-food chain over the past three decades. He describes a shift awayfrom the industrialized form of agriculture, a form which creates major sus-tainability problems. The shift is not yet completed but has progressed along way. Switzerland is one of the leading Western countries in sustainableagriculture, balancing economic, ecological and social dimensions. In thistransition two new forms of agriculture play a role. Organic farming takesa holistic stance, respecting the principles of nature, by seeking to maintainlong-term fertility and biological activity of soils using locally adapted bio-logical and mechanical methods as opposed to reliance on external inputs.Integrated production is a third way between organic farming and indus-trialized agriculture. In his analysis, Belz proposes a number of additionsto Geelss multi-level perspective.

In Chapter 6, Aad Correlj and Geert Verbong describe the transition tothe use of natural gas in the Netherlands in the 1950s and 1960s. The dis-covery of a large deposit of natural gas in 1959 caused a shock to the energysystem based on coal and coal-based gas. They distinguish three dimen-sions of the gas regime, notably (i) the material network, (ii) the institu-tional framework and (iii) the market for energy. They show that thetransition to a new system required a process of interrelated changes onthese three dimensions. They also show the transition cannot be under-stood without taking note of earlier developments that took place in the1950s.The authors analyse in detail the strategies, visions and activities ofrelevant actors, and show the struggles and negotiations that took place.

Part III: Transition Policy

In Chapter 7, Ren Kemp and Jan Rotmans present a general frameworkfor transition management. They argue that policy interventions shouldtarget not just economic conditions (through taxes and regulations) but alsobeliefs, expectations and institutional factors. They propose a managementstrategy based on modulation of ongoing dynamics rather than planningand control. The overall steering philosophy is to embark on a process oflearning-by-doing. This involves articulation of future visions, setting upexperiments to learn about the feasibility of visions, and the evaluation andadjustments of visions. Transition management is not a one-off exercise butinvolves several policy cycles of adjustment and learning. In that sense it isgoal-oriented incrementalism. The authors apply their ideas to the domainof transport and mobility. They show what the first step of transition man-agement might look like for a transition towards sustainable transport.

12 Introduction

In Chapter 8 Geert Teisman and Jurian Edelenbos argue that manage-ment of transitions requires a transition of management, that is, a transi-tion from hierarchical steering to interactive forms of governance. Thisrequires institutional change because, as they state it, the best way to kill anew idea is to put it in an old organization. The authors identify three bar-riers which hinder transition management: (i) missing links between inter-active processes and formal decision-making; (ii) fragmented departmentalstructures of governmental organizations frustrating productive andinnovative interactions; and (iii) the reluctance of public actors to shareresponsibility and accountability with each other and with private actors orsocietal actors. They discuss several experiments as possible forerunners fornew democratic governance systems: parallel democracy, hybrid democ-racy and participatory democracy. In current Dutch practice, the first formis advocated as a model for system innovation. The authors argue that thisis not sufficient and that a change should take place towards a hybriddemocracy or participatory democracy. This requires a redefinition of therole of various actors, especially at various levels of government.

Part IV: Tools for Transition Policy and Empirical Illustrations

In Chapter 9 Halina Brown, Philip Vergragt, Ken Green and Luca Berchiccidiscuss bounded socio-technical experiments (BSTE) as attempts to intro-duce a new technology, service, or social arrangement on a small scale.Based on insights from theories of organizational learning, policy-orientedlearning and diffusion of innovation, the authors identify two types oflearning: technical single-loop learning, and higher-order social learning.The first type of learning occurs among the participants in the experimentand their immediate professional networks. The second type occurs insociety at large. The authors argue that both types play a key role in a tran-sition towards sustainable mobility systems. They analyse two Dutchexperiments in personal mobility, the development of a three-wheeledbike-plus vehicle called Mitka and an attempt to solve mobility problemson the island of Texel. The cases show that the first type of learning tookplace to a considerable extent and that it can be facilitated by deploymentof structured visioning exercises, by diffusion of ideas among relatedBSTEs, by innovative couplings of problems and solutions, and by creatinglinks among related experiments. The cases also show that the second typeof learning was more difficult. The authors provide recommendations onhow both kinds of learning could be organized, stressing the importance ofvisions and vision-building processes.

In Chapter 10 Sirkku Kivisaari, Raimo Lovio and Erja Vyrynen take asa starting point that experimenting with alternatives to an existing system

Introduction 13

can play a crucial role in broader transition processes because they providethe seeds for change. They use the so-called societal embedding of inno-vations approach to analyse management of experiments. This approachhas been designed to enhance commercialization of innovations that yieldfinancial profit as well as contribute to sustainable development. It has beengeared especially towards supporting collaboration between public andprivate actors in cases where there is a considerable public interest infinding innovative solutions to societal issues. The chapter discusses twoFinnish experiments, which can be perceived as pilots of system innova-tions. The first deals with a new energy service company concept in Finnishmunicipalities and the second with development of a new diabetes self-management system. Combining findings from both cases, they discusshow the management of these experiments can be strengthened so thattheir results can indeed form the seeds for a transition.

In Chapter 11 Boelie Elzen, Frank Geels, Peter Hofman and Ken Greenpresent a new scenario method to explore future system innovations andsupport transition management. In a strict sense, transitions cannot besteered, because of their complex nature, but it is possible to stimulatedevelopments in more sustainable directions over a longer period of time.This requires a vision of which directions that might be, that is which com-bination(s) of technologies and their societal embedding might contributeto a sustainable system. To help develop such visions, scenario studies orother foresight methods can be used. Although many such methods exist,the authors argue that they have limitations for exploring system inno-vations and transitions. Hence, they present a new scenario method, calledSocio-Technical Scenarios. The chapter describes the main features of themethod and illustrates it by describing two short scenarios for the passen-ger mobility domain. The authors thus provide a concrete tool to helpdevelop guiding visions.

In Chapter 12 the editors take stock of the findings of the book, andsuggest a research agenda for the future.

REFERENCES

Abernathy, W.J., and K.B. Clark (1985), Innovation: mapping the winds of creativedestruction, Research Policy, 14, 322.

Arthur, Brian (1988), Competing technologies: an overview, in Giovanni Dosi,Chris Freeman, Richard Nelson, Gerald Silverberg and Luc Soete (eds),Technical Change and Economic Theory, London: Pinter, pp. 590607.

Berkhout, F. (2002), Technological regimes, path dependency and the environ-ment, Global Environmental Change, 12, 14.

14 Introduction

Breschi, Stefano, and Franco Malerba (1997), Sectoral innovation systems:technological regimes, Schumpeterian dynamics, and spatial boundaries, inCharles Edquist (ed.), Systems of Innovation: Technologies, Institutions andOrganizations, London and Washington, DC: Pinter, pp. 13056.

Callon, Michel (1991), Techno-economic networks and irreversibility, in JohnLaw (ed.), A Sociology of Monsters, Essays on Power, Technology and Domination,London: Routledge, pp. 13261.

Chandler, Alfred D. (2001), Inventing the Electronic Century: The Epic Story of theConsumer Electronics and Computer Industries, New York: The Free Press.

Christensen, Clay (1997), The Innovators Dilemma: When New Technologies CauseGreat Firms to Fail, Boston, MA: Harvard Business School Press.

Coombs, Rod, Ken Green, Albert Richards and Vivian Walsh (eds) (2001), Tech-nology and the Market: Demand, and Innovation, Cheltenham, UK: Edward Elgar.

Cowan, R. (1990), Nuclear power reactors: a study in technological lock-in,Journal of Economic History, 50, 54167.

Cowan, R., and P. Gunby (1996), Sprayed to death: path dependence, lock-in andpest control strategies, Economic Journal, 106, 52142.

David, P.A. (1985), Clio and the economics of QWERTY, American EconomicReview 75, 33237.

De Bruijn, Johan A., Walter J.M. Kickert and Johannes F.M. Koppenjan (1993),Inleiding: Beleidsnetwerken en overheidssturing, in Johannes F.M. Koppenjan,Johan A. De Bruijn and Willem J.M. Kickert (eds), Netwerkmanagement in hetopenbaar bestuur, The Hague: Vuga, pp. 1130.

Dosi, G. (1982), Technological paradigms and technological trajectories: a sug-gested interpretation of the determinants and directions of technical change,Research Policy, 6, 14762.

Du Gay, Paul, Stuart Hall, Linda Janes, Hugh MacKay and Keith Negus (1997),Doing Cultural Studies: The Story of the Sony Walkman, London: SagePublications.

Eggerton, D. (1999), From innovation to use: ten eclectic theses on the historio-graphy of technology, History and Technology, 16, 11136.

Etzkowitz, H., and L. Leydesdorff (2000), The dynamics of innovation: fromnational systems and Mode 2 to a triple helix of UniversityIndustryGovernment relations, Research Policy, 29(2), 10923.

Geels, F.W. (2002), Technological transitions as evolutionary reconfiguration pro-cesses: A multi-level perspective and a case-study, Research Policy, 31, 125774.

Hughes, Thomas P. (1994), Technological momentum, in Merrit R. Smith and LeoMarx (eds), Does Technology Drive History? The Dilemma of TechnologicalDeterminism, Cambridge, MA: The MIT Press, pp. 10113.

Jacobbson, S., and A. Johnson (2000), The diffusion of renewable energy technology:an analytical framework and key issues for research, Energy Policy, 28, 62540.

Kline, R., and T. Pinch (1996), Users as agents of technological change: the socialconstruction of the automobile in the rural United States, Technology andCulture, 37, 76395.

Kohler-Koch, Beate, and Rainer Eising (eds) (2000), The Transformation ofGovernance in the European Union, London and New York: Routledge.

Kooiman, Jan (ed.) (1993), Modern Governance: New GovernmentSocietyInteractions, London: Sage.

Leonard-Barton, Dorothy (1995), Wellsprings of Knowledge: Building andSustaining the Sources of Innovation, Boston, MA: Harvard Business School Press.

Introduction 15

Lie, Merete, and Knut H. Srensen (eds) (1996), Making Technology Our Own:Domesticating Technology into Everyday Life, Oslo: Scandinavian UniversityPress.

Lynn, L.H., N.M. Reddy and J.D. Aram (1996), Linking technology and institu-tions: The innovation community framework, Research Policy, 25, 91106.

Malerba, F. (2002), Sectoral systems of innovation, Research Policy, 31(2), 24764.National Research Council (1999), Our Common Journey. A Transition Toward

Sustainability, Washington, DC: National Academy Press.Nelson, Richard R., and Sidney G. Winter (1982), An Evolutionary Theory of

Economic Change, Cambridge (MA): Bellknap Press.Oudshoorn, Nelly, and Trevor Pinch (eds) (2003), How Users Matter: The Co-

Construction of Users and Technology, Cambridge, MA: MIT Press.Raskin, Paul, Tariq Banuri, Gilberto Gallopin, Pablo Butman, Al Hammond,

Robert Kates and Rob Swart (2002), Great Transition. The Promise and Lure ofthe Times Ahead, Boston: Stockholm Environment Institute and Global ScenarioGroup.

Rhodes, Ron A.W. (1997), Understanding Governance: Policy Networks, Govern-ance, Reflexivity and Accountability, Buckingham: Open University Press.

Schot, J., R. Hoogma and B. Elzen (1994), Strategies for shifting technologicalsystems. The case of the automobile system, Futures, 26, 106076.

Schwartz-Cowan, Ruth (1987), The consumption junction: a proposal for researchstrategies in the sociology of technology, in Wiebe E. Bijker, P. Hughes Thomasand Trevor Pinch (eds), The Social Construction of Technological Systems: NewDirections in the Sociology and History of Technology, Cambridge, MA: The MITPress, pp. 26180.

Stankiewicz, Robert (1992), Technology as an autonomous socio-cognitivesystem, in Harold Grupp (ed.), Dynamics of Science-Based Innovation, Berlin:Springer-Verlag, pp.1944.

Tushman, M., and P. Anderson (1986), Technological discontinuities and organ-ization environments, Administrative Science Quarterly, 31, 49365.

Unruh, G.C. (2000), Understanding carbon lock-in, Energy Policy, 28, 81730.Van de Ven, Andrew H., and Raghu Garud (1989), A framework for understand-

ing the emergence of new industries, in Richard Rosenbloom and RobertBurgelman (eds), Research on Technological Innovation, Management and Policy,Greenwich, CT: JAI Press, pp. 195225.

Van Heffen, Oscar, Walter J.M. Kickert and Jacques J.A. Thomassen (eds) (2000),Governance in Modern Societies: Effects, Change and Formation of GovernmentInstitutions, Dordrecht: Kluwer Academic Publishers.

Vellinga, Pier, and Nadia Herb (eds) (1999), Industrial transformation scienceplan, International Human Dimensions Programme report no. 12.

VROM (2001), Een wereld en een wil, werken aan duurzaamheid, NationaalMilieubeleidsplan 4, Dutch Environmental Policy Plan 4, established by nineDutch departments under the coordination of VROM (The Department ofHousing, Spatial Planning and Environment), The Hague: VROM.

Walker, W. (2000), Entrapment in large technology systems: institutional commit-ments and power relations, Research Policy, 29, 83346.

Weterings, Rob, Jan Kuijper, Erik Smeets, Gert J. Annokke and Bert Minne (1997),81 Mogelijkheden, Technologie voor Duurzame Ontwikkeling. Eindrapport van demilieugerichte technologieverkenning, The Hague: VROM (Department ofHousing, Spatial Planning and Environment).

16 Introduction

PART I

Theoretical explorations of transitions

2. Understanding system innovations:a critical literature review and aconceptual synthesisFrank W. Geels

INTRODUCTION

System innovations are defined as large-scale transformations in theway societal functions such as transportation, communication, housing,feeding, are fulfilled. Technology plays an important role in fulfilling soci-etal functions. Artefacts by themselves have no power, they do nothing. Onlyin association with human agency and social structures and organizationsdo artefacts fulfil functions. In real-life situations (for example, organiza-tions, firms, houses) we never encounter artefacts per se, but artefacts-in-context. For the analysis of working/functioning artefacts in context, it is thecombination of the social and the technical that is the appropriate unit ofanalysis (Fleck, 1993, 2000). From the perspective of science and technologystudies two basic notions of technology are important: (i) technology is het-erogeneous, not just a material contraption (engineers know this, their workis heterogeneous engineering); (ii) the functioning of technologies involveslinkages between heterogeneous elements. Hughes (1987) coined themetaphor of a seamless web to indicate how physical artefacts, organizations(for example, manufacturing firms, investment banks, research and devel-opment laboratories), natural resources, scientific elements (that is, books,articles), legislative artefacts (laws) are combined in order to achieve func-tionalities. From these considerations it follows that societal functions arefulfilled by socio-technical systems. Socio-technical systems consist of acluster of elements, including technology, regulation, user practices andmarkets, cultural meaning, infrastructure, maintenance networks, supplynetworks (see Figure 2.1 for an example for land-based transportation).

In this conceptualization, a system innovation can be understood asa change from one socio-technical system to another. One aspect ofa system innovation is technological substitution, which comprises threesub-processes: (i) emergence of new technologies, (ii) diffusion of new

19

technologies, (iii) replacement of old by new technology. The second aspectis coevolution. System innovations not only involve technological substitu-tions, but also changes in elements such as user practices, regulation, indus-trial networks, infrastructure, and cultural meaning. The third aspect is theemergence of new functionalities. When radical innovations have particulartechnical properties, this may enable the articulation of new functionalcharacteristics. Radical innovations may then introduce new functionalitiesand change the way in which performance is measured (Abernathy andClark, 1985; Utterback, 1994; Christensen, 1997).

This chapter addresses the following question: how do major changes insocio-technical systems occur?

LITERATURE REVIEW AND CRITICALEVALUATION

There are few literatures which discuss all aspects of system innovations.Literatures typically focus on one or two aspects, but make simplisticassumptions about other aspects. Thus literatures provide bits and pieceswhich can be used as building blocks for a more integrative perspective.I take technological substitution as the entry point for the literature review,and discuss the other aspects when I discuss existing literatures. In theliteratures there are large differences about what it is that changes duringtechnological substitution, and the kind of change process. I will de-scribe and critically evaluate a range of sociological, economic and socio-technical literatures. I distinguish three basic approaches on systeminnovations in these literatures: point-source approaches, replacement

20 Theoretical explorations of transitions

Socio-technical systemfor transportation

Culture and symbolicmeaning (e.g. Freedom, individuality )

Regulations and policies(e.g. traffic rules, parking fees,emission standards, car tax)

Road infrastructureand traffic system(e.g. lights, signs)

Vehicle (artefact )

Markets and user practices(mobility patterns, driver preferences )

Industry structure (e.g. car manufacturers,suppliers )

Maintenance and distribution network (e.g. repair shops, dealers)

Fuel infrastructure (oil companies, petrol stations )

Figure 2.1 Socio-technical system for modern car-based transportation

approaches, and transformation approaches. Point-source approaches typi-cally focus on emergence and diffusion of (radical) novelties, but say littleabout replacement. Replacement approaches focus on the competitionbetween old and new technologies, but typically make simplistic assump-tions about the emergence and diffusion of novelties. Transformationapproaches focus mainly on emergence of novelties, and how rules and per-ceptions of novelties gradually change. Each of these approaches has inter-esting insights and weaknesses.

Point-source Approaches

Literatures in these approaches focus on the emergence and diffusion of newtechnologies but do not say much about replacement. Change begins as apoint source, initiated by the emergence of a (radical) novelty. Subsequently,the novelty conquers the world.

Technology life cycle approachIn technology life cycle approaches several phases are distinguished. Themain focus is on the market shares of technologies, firm strategies andmarket structures.

In the first phase (birth, childhood) a new technology is born. The newtechnology exists as a variety of products, and has low production volumesand market shares. There is technological uncertainty, and uncertaintyabout user preferences. Learning processes are targeted towards productinnovation. The industry structure is fluid, consisting of many small firms,and high rates of entry and exit.

In the second phase (adolescence), the initial diversity gives way to stand-ardization, leading to a dominant design. The rate of product innovationslows down. Process innovations become more important to lower costs,and conquer higher market shares. Concentration and shake-out occur andthe industry structure stabilizes.

In the third phase (maturity) growth rates slow down as markets becomesaturated and improvements face diminishing returns. The market becomesconcentrated in the hands of a few producers, leading to an oligopolisticindustry structure. Producers try to squeeze out the last marginal costimprovements from scale economies.

Economic path-dependency theoriesIn economic path-dependency theories the market share of technologies isalso what changes. The focus is on self-reinforcing processes, leading toincreasing returns of adoptions (David, 1985; Arthur, 1988). This meansthat the more a particular technology is used, the greater its attractiveness

Understanding system innovations 21

relative to its competitors. Arthur (1988: 591) identified five sources ofincreasing returns to adoption (IRA): (i) learning by using: the more a tech-nology is used, the more is learned about it, the more it is improved; (ii)network externalities: the more a technology is used by other users, thelarger the availability and variety of (related) products that become avail-able and are adapted to the product; (iii) scale economies in production,allowing the price per unit to go down; (iv) informational increasingreturns: the more a technology is used, the more attention it receives, stim-ulating others to adopt; (v) technological interrelatedness: the more a tech-nology is used, the more complementary technologies are developed.

Although these five mechanisms of IRA are certainly relevant for thediffusion of new technologies, they do not say much about emergence andreplacement. Path dependency literature is unconcerned with questionsrelating either to the existence of prior technologies, or to the way in whichnew technologies are able to displace older technologies.

Science and technology studies: SCOT, ANT and LTSAn influential approach in science and technology studies is SCOT (SocialConstruction of Technology). In the early, SCOT1 approach (Pinch andBijker, 1987; Kline and Pinch, 1996) the focus is on sociocognitiveprocesses (meaning and interpretation in social groups). The main aim isto understand the form and function of new technologies. Why do newtechnologies stabilize into a particular form, and how are they used? Toanswer this question, the focus is on the relevant social groups which areinvolved in the development process, such as engineers, users, policymakers, societal groups, and so on. These social groups may have differentideas about problems, solutions and meanings of the artefact. There isinterpretative flexibility. Gradually a consensus emerges about the domi-nant meaning of an artefact, leading to stabilization. Selection is thus seenas a sociocognitive process (closure and stabilization of one interpretationin social groups). The later SCOT2 approach (for example, Bijker, 1995)gave the analysis a more structuralist flavour by adding the notion of atechnological frame to look at structural social, cognitive and material ele-ments. The technological frame comprises elements such as: goals, keyproblems, problem-solving strategies (heuristics), requirements to be metby solutions to problems, current theories, tacit knowledge, testing proce-dures, and design methods and criteria. The SCOT analysis stops whenthe artefact, social groups and technological frame have stabilized. SCOTdoes not say much about wider diffusion of new technologies. Thereplacement of old technologies is not dealt with in the framework.

In the research approaches of Large Technical Systems (LTS) and actor-network theory (ANT) the focus is on linkages in and around emerging

22 Theoretical explorations of transitions

technologies. In both perspectives the dynamic is that heterogeneous ele-ments are gradually linked together, emphasizing coevolution.

In LTS research the emergence and development of large technicalsystems is loosely described as life-cycle (Hughes, 1983, 1987; Mayntzand Hughes, 1988). Several types of system builder (inventor, inventor-entrepreneur, manager-entrepreneurs, financier-entrepreneurs) are activein different phases: invention, development, innovation, growth, competi-tion and consolidation, momentum. System builders like Thomas Edisonare heterogeneous engineers, working not only on physical materials, butalso on people, texts, devices, city councils, economics and so on. Hughescoined the term seamless web to indicate the heterogeneous character ofLTS. In the early phases, the web and its linkages were fragile, requiringEdison to put in a lot of work to uphold it. As the electricity network grewand stabilized, it gained momentum and began to have coordinatingeffects. A reversal occured as the technology shifted from flexibility todynamic rigidity (Staudenmaier, 1989). The technology shifts roles froma possible social option to a culture-shaping and highly specified socialforce.

The perspective of socio-technical linkages is most consistently developedin ANT (Latour, 1987, 1991, 1992, 1993; Callon, Law and Rip, 1986; Callon,1991; Callon et al., 1992). New technologies emerge from the start asheterogeneous configurations. In the early phase of a new technology, thenetwork consists of few elements and linkages. Innovation is about theaccumulation of elements and linking them together in a working configur-ation. As the network is expanded and more elements are tied together, atechnology becomes more real. Diffusion is also a process of creatingsocio-technical linkages. The diffusion of an artefact across time and spaceneeds to be accompanied by an expansion of linkages in which the artefactcan function, such as, test apparatus, spare parts, maintenance networks,and infrastructure. Thousands of people are at work, hundreds of thou-sands of new actors are mobilized (Latour, 1987: 135).

These different point-source literatures are useful because they distin-guish patterns or phases in the emergence and diffusion of new technologies.They do not say much, however, about technological replacement. Becauseof their focus on new technologies, they tend to neglect the existence of oldtechnologies.

Replacement Approaches

Literatures in these approaches focus mainly on economic competition andsubstitution. The focus is on technologies which compete with each otherin markets on the basis of cost and performance. The emergence of new

Understanding system innovations 23

technologies is often conceptualized in a simplistic way, for example as astochastic process, driven by individual genius.

Technological and economic substitution approachesTechnological and economic substitution approaches understand replace-ment as a market-based process, in which new technologies replace incum-bent technologies, because of higher performance and lower price.

Grbler (1991, 1998) and Nakicenovic (1986, 1991) have initiated a par-ticular approach to long-term technological replacements. Their basicassumption is that the replacement of an old technology by a new tech-nology proceeds along the logistic substitution curve: f/(1f) e (a.t b),in which t is the independent variable representing some unit of time, aand b are constants, f is the fractional market share of the new competi-tor and (1f) is the fractional market share of the old one. Logistic curvemodels are entirely descriptive, and do not explain why curves behave asthey do.

The competitive dynamic is made more explicit in neo-classical economicapproaches. Buyers compare price and performance of rival technologies.Users are represented as having a fixed set of user preferences. A new tech-nology replaces an old technology, if its performance characteristics have abetter fit with the user preferences and users buy more of it. The user is rep-resented as a rational actor, who has some kind of formula in his head tomake optimal adoption choices.

To describe developments in price and performance over time, the conceptof learning curves was developed. The performance of technologies in-creases as organizations and individuals gain experience with them, that is,organizational and individual learning by doing and learning by using(Arrow, 1962; Rosenberg, 1982). Learning depends on the actual accumula-tion of experience. Learning curves are generally described in the form of apower function, measured as cumulative output: Y a Xb , where y is thecost or performance of the xth unit, a is the cost associated with the first unit,and b is a parameter measuring the cost reductions for each doubling ofcumulative output (that is, the learning rate). If the learning rate of new tech-nologies is higher than that of established technologies then the former willeventually replace the latter.

A first criticism is that replacement approaches assume that new tech-nologies compete in the same markets as the old technologies. This assump-tion can be questioned with regard to the early phases of new technologies.The first automobiles did not compete with horse-and-carriages. Insteadthey were used for pleasure and adventure, in racing and touring. Similarly,the first steamships did not compete with merchant sailing ships, but wereused for auxiliary functions, such a tow boats or pirate hunters (Geels,

24 Theoretical explorations of transitions

2002a). More generally, substitution approaches are unclear about theemergence of new technologies.

A second criticism is that substitution approaches suggest that old andnew technologies always have a relationship of competitive struggle.Although competition certainly plays a role in the life cycle of new tech-nologies, it is not necessarily true for the early phases. Railroads did notimmediately compete with canals and water transportation, but were usedas feeders to them (Rosenberg, 1976: 197). New technologies may also formtechnical hybridizations with old technologies. Steam engines were firstused as additional power sources on sailing ships to be used when there wasno wind.

Third, the conceptualization of the demand side is static (fixed user pref-erences). Although user preferences may be assumed relatively stable in theshort-run, they definitely change over longer time periods. Particularly withvery new technologies, users may develop new preferences, practices andnew cognitive categories. The evolution of consumer preferences is anunderdeveloped area in economics, but is recently being taken up in inno-vation studies and technology studies. Consumption acts are nested intocognitive categories and mental models of the actors (Aversi et al., 1999).Furthermore, adoption is not a passive act, because a product has to beintegrated in user practices. These domestication processes may involveinnovations in organization routines, work practices, management styles,symbolic meaning (Lie and Srensen, 1996).

Punctuated equilibria and technology cycles approachIn technology management and industrial economics the concepts of punc-tuated equilibria and technology cycles have been coined (Tushman andAnderson, 1986; Anderson and Tushman, 1990; Rosenkopf and Tushman,1994; Tushman and Murmann, 1998). It is argued that technological devel-opment constitutes an evolutionary process punctuated by discontinuouschange. For long periods of time technological change is relatively stable,proceeding incrementallydowndesignhierarchiesandtechnical trajectories.These periods of incremental change are punctuated by brief periods ofrapid change. An era of ferment is triggered by the emergence of a techno-logical breakthrough, which is relatively rare and tends to be driven by indi-vidual genius(Tushman and Anderson, 1986: 440). Because a revolutionaryinnovation is crude, different design options are tried, creating uncertaintyabout which design will win. The period of ferment is closed by the emer-gence of a dominant design. A period of incremental technical change thenfollows, until it is broken by the next technological discontinuity.

One problem is that the punctuated-equilibrium analysis is implicit