Technological Forecasting & Social Change 72 (2005) 651 – 661

Introduction

Transitions towards sustainability through system innovation

Boelie Elzena, Anna Wieczorekb,T
a
Science, Technology and Society, University of Twente, FWT/BBT-Het Capitool, P.O. Box 217,
7500 AE Enschede, The Netherlands
b
Institute for Environmental Studies (IVM), Vrije Universiteit in Amsterdam, de Boelelaan 1087,
1081 HV Amsterdam, The Netherlands
Received 31 March 2005; accepted 20 April 2005

1. Introduction

The road towards sustainability poses many challenges. There are many areas of human needs to be
dealt with such as food, fresh water, health, shelter, mobility, energy, and many others. On top, there are
many dimensions on which sustainability needs to be achieved, including technical, socio-economic,
cultural, spatial, environmental preservation, distribution of wealth, etc.
Achieving sustainability in the broad sense therefore requires a dazzling multitude of changes. One of
the approaches suggests that an encompassing set of transformations will be needed that have been
referred to by analysts from different disciplinary backgrounds using a variety of concepts. Some of the
best known include dsystem innovationT, dregime transformationT, dindustrial transformationT,
dtechnological transitionT, dsocio-economic paradigm shiftT. We use the shorter phrase dtransitionT to
cover all of these in this article.
Loosely defined, a transition denotes a long-term change in an encompassing system that serves a
basic societal function (e.g. food production and consumption, mobility, energy supply and use,
communication, etc.). In a transition, both the technical as well as the social/cultural dimensions of such
a system change drastically. This emphasis on the co-evolution of technical and societal change
distinguishes transitions from incremental processes, which are primarily characterised by technical
change (through successive generations of technologies) with relatively little alteration of the societal
embedding of these technologies.

T Corresponding author.
E-mail addresses: b.elzen@utwente.nl (B. Elzen)8 anna.j.wieczorek@ivm.vu.nl (A. Wieczorek).

0040-1625/$ - see front matter D 2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.techfore.2005.04.002
652 B. Elzen, A. Wieczorek / Technological Forecasting & Social Change 72 (2005) 651–661

The notion of transition has increasingly gained attention over the past years, in academic as well as in
policy arenas. Policy makers are especially interested in transitions since incremental change is not
believed to lead to sustainability. Transition is perceived as a policy objective, which has a great potential
to tackle current problems in various life domains.
On the academic side, several workshops and conference sessions have been organised on these
themes in the past years. To name a few: a first international meeting was the workshop on dTransitions
towards sustainability through system innovationT, held at the University of Twente in the Netherlands in
de summer of 2002. Several papers presented at this workshop have recently been published in an edited
volume [1]. Another important event was the 2003 dOpen Meeting of the Human Dimensions of Global
Environmental Change Research CommunityT held in Montreal, Canada. The authors of this
introduction organised several sessions on transitions at this conference and this special issue collects
some of the session contributions.
The current research and up to date experience with transitions suggest that to be able to stimulate
transitions towards sustainability we still need to build a strong knowledge base in this field. In particular
we face two major research challenges, notably (1) the need to improve insights in the dynamic of
transition processes and (2) the need to successfully apply these insights in the development of strategies
and policies that would induce and stimulate the occurrence of transitions. In this introductory article, we
will briefly discuss some of the main elements of these two challenges. The first section addresses the
issue of dunderstanding transitionsT while the second addresses the topic of dinducing transitionsT. In this
discussion we will also introduce the various papers presented in this special issue and their contribution
to each of the research challenges.

2. Understanding transitions

In the 1960s and 1970s, it became widely acknowledged that the development of industrial societies
has had serious negative effects on the life support systems. Production and consumption patterns caused
large-scale pollution and the destruction of the natural environment. While environmental policies have
succeeded in addressing many of the most urgent (often of local and regional character) problems, they
fell short in tackling persistent global environmental problems. An example of such a problem is climate
change caused by excessive CO2 emissions originating, to a great extent, from human activities,
especially the massive use of fossil fuels. The activities, driven by continued economic growth and
increasing levels of consumption are global in scope but deeply embedded in our cultures, which is one
of the reasons that makes global environmental problems so difficult to approach. Even despite the
adoption of the notion of sustainable development by most governments as a basic policy principle, it is
clear that the transformation to a dpost-industrialT society will not necessarily and simultaneously be a
transition towards sustainability, i.e. a society that is characterised by a better balance between economic,
social and ecological performance. Ensuring that a transition does lead to more sustainability became a
major challenge for society in general and for environmental policy in particular.
Over the past decade the transition to sustainability has become a central theme in the work of many
organisations, including government bodies, NGOs, professional organisations, researchers, global
environmental change networks, etc. Technological change has been defined as crucial for such a
transition by various national governments and in various publications, including recent White Papers by
the EU [2]. This position, however, has met fierce criticism by those who portray technology as part of the
 B. Elzen, A. Wieczorek / Technological Forecasting & Social Change 72 (2005) 651–661 653

problem and emphasise the need for changes in the culture and structure of our current highly consumptive
economies, via e.g. more sustainable consumption patterns and habits. Both positions are too simple,
however, because technological change cannot be so easily separated from structural and cultural change.
The key starting point in this article and the others in this special issue is that both kinds of changes are
highly interrelated and mutually dependent on each other, forming a sort of a seamless web. For example,
any change towards more sustainable mobility patterns will require new technologies interacting with
social and cultural changes. The consequence is that any transition to sustainability will imply a
substantial degree of socio-cultural change coupled to a similar high level of technological change.
Thus, transitions are extremely complicated processes, which makes transition research an even more
complex type of endeavour. In our attempt to propose some structure in this field, we will start by briefly
discussing transition processes as innovation processes. Starting from this angle, we can make a
distinction between radical and incremental forms of innovation to help clarify the notion of transition.
Innovation studies have shown that most innovation tends to be of an incremental nature. New
technologies developed and brought to market are typically variants of existing technologies that can be
used by customers with little or no extra instruction or training, using existing infrastructures and
allowing them to use it in a way they always did, possibly with some extra features valued by the user.
With successive generations of technology a pattern of gradual technical change thus emerges with
relatively little change in terms of the societal embedding of this technology. Innovation studies use the
concept of dtechnological trajectoryT to describe such a path of development [3].
At the same time, radical alternatives to such a trajectory are developed but because of a variety of
reasons they have a hard time breaking through. They face huge barriers in the existing situation, barriers
related to production processes (with high knowledge and capital investments), regulation, user
preferences, infrastructures, etc. With such barriers it is no surprise that most innovations that gain a
substantial market share tend to be of an incremental nature.
Going back to the issue of sustainability, the distinction between incremental innovation and transition
helps to clarify the discrepancy between what is actually happening and what seems to be needed. What
has happened over the past decades is that attempts to achieve various societal objectives such as a
cleaner environment have resulted in a wide variety of incremental changes. Some problems, however,
appear to be very persistent and more radical changes seem necessary to tackle these.
To take an example: in the domain of passenger mobility, the incremental path has rendered
significant successes. Since the 1960s, emissions of atmospheric pollutants have decreased by a factor of
10 or more. Fuel efficiency of cars has improved considerably. With already proven technologies, both
characteristics can be improved further. These achievements may seem promising but a nagging question
remains, which is whether the continuation of this incremental path will lead to a sustainable transport
system in the broad sense. There are good reasons to doubt that this will happen. Internationally, the
issue of CO2 emissions is presently high on the environmental agenda. In the longer term, reductions of
80% or more are thought to be necessary. There is general consensus that this cannot be achieved only
by improving the efficiency of vehicles with internal combustion engines. Automakers as well as many
others with technical blinkers then point to the potential of fuel cells in achieving the required reductions.
The future may prove they were right, but it might as well not. The careful conclusion should then be
that, it is recommendable to explore the hydrogen option further but it is (also) risky to bet on it.
Furthermore, sustainable transport is not just an issue of low emissions of pollutants and CO2. It is
also an issue of accessibility and a more general concept of liveable cities. To tackle these challenges a
more drastic change of the current transport system seems required. This system is characterised by the
654 B. Elzen, A. Wieczorek / Technological Forecasting & Social Change 72 (2005) 651–661

massive use of cars that are mostly used to transport just one person for an average of 1 h/day. The rest of
the time cars just sit idle and use precious urban space, which gives the car system an overall efficiency
of about 1%.1 Only a drastic improvement of this efficiency would render a sustainable system, which
implies that the role of the car would have to be played down and replaced with other means of
transportation. Given this broader sustainability challenge the incremental innovation might not suffice
and we may find ourselves in need for a more radical innovation. This radical path would seek to
develop new vehicle technologies (mass transport or the other extreme, vehicles tuned to transport just
one passenger), new infrastructures, new logistics, new legislation and, not least important, change of
travel behaviour—a true transition indeed.
The required transition poses an enormous challenge, obviously too large for current generations of
policy makers and transport planners. The typical approach is to improve the current system by small
incremental steps, to tackle the most pressing problems and not worry about where such an approach
might eventually lead. The car continues to be treated as a holy cow and more innovative minds, who
argue that more fundamental changes are needed and even possible, meet general scepticism.
In conclusion, a transition in the mobility system seems to be needed while, at the same time, it
appears extremely difficult if not impossible to achieve. A comparable argument can be made for many
other areas of basic human needs, especially if we look at issues like the supply of fresh water, food
production, energy production and consumption at a global scale.
The main challenge for research then is to gain a better understanding of the dynamics of transitions in
such a way that it also suggests possibilities to induce and stimulate the occurrence of transitions. As a
first step, let us try to be a bit more specific on the nature of transitions. One main characteristic follows
from the ambition to make these insights of use for sustainability challenges. Many sustainability
problems are defined at the level of some of the basic human needs mentioned before, including water,
food, energy, mobility, etc. It seems sensible to define transitions at the same level, i.e. the level of the
regimes (or systems)2 related to these needs, notably the water regime, food regime, mobility regime,
etc. This leads to the first main characteristic of transitions:
Characteristic 1: Transitions are defined to occur in encompassing regimes (systems) in relation to
basic human needs. Each of these regimes is characterised by a range of technologies,
infrastructures, patterns of behaviour, cultural values, policies, etc.
The second main characteristic comes from the above discussion on innovation processes and the
distinction between incremental change and a transition. It can be described as follows:
Characteristic 2: Transitions imply change processes that affect all or a large part of the dimensions
mentioned in the first characteristic. They are at least characterised by a combination of technical
and societal/behavioural change, also described as a process of dco-evolutionT.
These characteristics are clearly underpinned in the first article that follows in this issue, by Green
and Foster [8]. They analyse a specific case in the domain of food production and consumption,

1
Vehicles being used about 4% of the time up to about one-quarter of their passenger capacity renders an efficiency of 1%.
2
In the literature, both concepts are used to indicate more or less the same entities. We prefer the concept of regime to system since most
studies using the latter usually emphasise technical characteristics, sometimes by explicitly calling them dLarge Technical SystemsT [4,5]. The
notion of regime stresses co-evolution, i.e. the interplay between technical and social change [6,7]. In this paper we will therefore use the
concept of regime, occasionally followed by a bracketed (system) as a service to readers more familiar with the latter term.
 B. Elzen, A. Wieczorek / Technological Forecasting & Social Change 72 (2005) 651–661 655

notably the dpea systemT in the UK. Like many other food products, peas are grown and processed in
an dindustrialT agricultural practice based on advanced breeding techniques and major inputs of
chemical fertilisers and pesticides. This mode of producing food is not very sustainable, also because
it requires high-energy processing. For this reason, many argue that such industrialised systems should
be dismantled and replaced with alternative methods of agriculture, food processing and distribution
that emphasise social and environmental sustainability. Green and Foster discuss the environmental
and social sustainability of different strategies for food systems by analysing the whole chain of
production, processing, distribution and consumption of the frozen peas. They focus attention on the
intimate connection between consumption and production and the way this connection renders
particular technological practices across the system. To achieve sustainability, all these practices
should be transformed.
Green and Foster thus clearly demonstrate that a transition to sustainability in this area would require
a radical change, i.e. a transition of the food system. Such a transition will involve the adoption and
diffusion of new technologies embedded in new economic, social, institutional and cultural relations.
Changes are required at the level of systems of production, distribution as well as consumption patterns
such as eating habits.
Going beyond their individual case study and building upon a variety of others3 we infer that
transitions, next to the two main characteristics discussed above, display the following specific
attributes:

! Multi-actor: They involve a wide range of actors, including firms, consumers, NGOs, knowledge
producers and governments;
! Multi-factor: They are not caused by a change in a single factor but are the result of the interplay of
many factors that influence each other. They are a combination of technical, regulatory, societal, and
behavioural change.
! Multi-level: They imply change at various levels: at the micro-level of individual actions, at the
meso-level of structuring paradigms and rules (regimes or systems) and at the macro-level
comprising wider societal and cultural characteristics and trends such as individualisation and
globalisation.

These specific attributes demonstrate that transitions are extremely complex processes, which
makes one wonder whether it is possible at all to comprehend and analyse them in a single
conceptual framework. In this issue, Geels [10] gives a positive answer to this question by proposing
what he calls the dmulti-level perspective on transitionsT. He conceptualises transitions as system
innovations, i.e. as changes from one socio-technical system to another. He distinguishes three
different levels. The meso-level of a socio-technical regime describes a specific system that the
analyst is interested in, e.g. the mobility regime. The macro-level is called the socio-technical
landscape, which describes broad factors that influence a variety of regimes, e.g. the problem of CO2
emissions and global warming (which may impact upon regimes like transport, energy, food
production, etc.). The micro-level is that of technological niches which is where radical innovations
emerge. The key feature is that system innovations occur through the interplay between the dynamics

3
Cf. various articles in [1] and in [9].
656 B. Elzen, A. Wieczorek / Technological Forecasting & Social Change 72 (2005) 651–661

at these multiple levels. In his article Geels shows that this analytical framework can account for the
stability in a regime (leading to incremental innovations as discussed above) as well as for
instabilities that can eventually lead to a transition.

3. Inducing transitions

Once we can understand and analyse transition processes the follow-on question is whether it is
possible to induce and sustain them in order to achieve sustainability. As mentioned above, a variety of
barriers make it difficult for novelties (or radical innovations) to break through to an existing regime (or
system). These barriers include production barriers, regulatory barriers, user preferences tuned to
existing regime, infrastructure requirements, investment needs, technological lock-ins, etc. Innovation
analysis to investigate these barriers tends to focus on the production side, i.e. on the role of innovators
and industries, either as whole sectors or as separate companies or even individuals. It is evident,
however, that public authorities also play an important role in innovation by either stimulating or
impeding it through various regulations, either intentionally or unintentionally.
Concerning the latter, existing legislation often constitutes a barrier for a transition since the
regulations are usually tuned to an existing regime with specific characteristics. As a result it may be
problematic or even illegal to work with a novelty in the public domain (e.g. with the first electric
vehicles). But there is also another side. Regulators can use their legislative powers to stimulate the
development and use of novelties. The same electric vehicles are a clear case in point as their use was
stimulated by various city and national governments. Regulators thus have a choice to either be very
active in stimulating innovation or to impede it. Which role they choose is influenced by their
assessment of the range of barriers listed above. If they have low expectations of the potential of the
novelty they are not likely to stimulate its development, if they have high expectations, they will try to
stimulate it.
Public authorities have a range of instruments at their disposal to influence the innovation process;
both in terms of the characteristics of technologies (e.g. through environmental standards) as well as
to either encourage or discourage their use (e.g. through taxes, subsidies). Which (combination of)
instruments they choose has been investigated widely in policy and governance studies but a major
shortcoming of this approach is that this is typically done from the perspective of realising a
relatively well defined short-term objective. Transitions, however, are about more vaguely defined,
distant objectives and longer-term processes. This therefore requires a re-assessment of the role of
public authorities.
The article by Taylor et al. [11] in this issue takes a detailed look at how government actions may
induce innovation. They use a combination of quantitative and qualitative analysis to investigate the case
of sulphur dioxide control technology for power plants. Distinguishing between dtechnology-pushT and
ddemand-pullT instruments they conclude that the former tend to be less effective in prompting invention.
But not only do governments affect innovation through direct regulation, they can and do also play a role
as mediator and facilitator, e.g. by fostering knowledge transfer via technical conferences, as well as
affecting the pattern of collaborative relationships within the technical research community via
regulatory changes that affect the market for the technology. Their analysis also implies a warning:
contrary to the common belief from many, they find little evidence for the claim that cap-and-trade
instruments induce innovation more effectively than other instruments.
 B. Elzen, A. Wieczorek / Technological Forecasting & Social Change 72 (2005) 651–661 657

Although this article teaches us important lessons on the (possible) role of government in innovation,
an important question remains regarding the relevance of these findings for the sake of inducing
transitions. Above, we made a distinction between incremental innovation and radical innovation and the
sulphur dioxide control case is clearly an example of the former. Let us try to take a step beyond this
specific case and discuss innovation policies in a somewhat broader perspective.
In recent literature some ideas have been proposed on how to distinguish incremental innovation
policies from policies seeking to induce transitions. The latter are often referred to as dtransition policyT
or dtransition managementT [12,13]. The complexity of transition processes and their long duration
implies a warning that such policies cannot be based on simple steering philosophies. Such policies will
need to take into account interaction between different stakeholders, unpredictability, a need to monitor
ongoing processes and adapt strategies if needed, and, by implication, a need to leave room for learning
and feedback [14]. Possibly, such policies will need to combine existing policy instruments with some
new approaches and assessment methods to identify the optimal mix in specific circumstances.
As stated above, transitions are multi-actor, multi-factor and multi-level and can only be understood in
terms of co-evolutionary processes that link up these actors, factors and levels. These processes are
characterised by high levels of uncertainty, unpredictability, and risk, and, therefore, are open-ended
learning processes. Influencing, rather than steering, such processes will certainly be difficult, but it may
not be impossible. Part of the challenge is to influence developments at an early stage, when
irreversibilities have not yet set in and one can still attempt to tip the balance between desirable and
undesirable developments.
The emphasis on learning and anticipation, however necessary, is not a guarantee that the desired
directions will be realised because attempts to set directions will be contested. Policy interventions need
to take the existing dynamic into account and cannot realise a drastic deviation from current courses of
development. The policy objective should therefore be to find ways to modulate the on-going dynamics
so that it bends slightly in the direction of desired objectives. A slight initial bend, however, provided the
new course is consistently maintained, can lead to drastically different outcomes in the longer term,
which, after all, is what transitions are about.
Although long-term visioning is not foreign to the political processes, sustainability transition policy
thinking is quite a new development. In recent years various scholars have taken up the challenge to
identify specific elements of policies seeking to induce transitions, identifying promises as well as
limitations [13,15]. These scholars however, come from different disciplinary backgrounds and for that
reason their approaches towards transition policy differ. At this stage of research it is not
counterproductive, still it is important that such ideas are related to each other and discussed in an
interdisciplinary fashion for the sake of identification of some common themes that can serve as a
stepping stone for more focused further work.
If we want to use research findings to develop more precise recommendations for transition policy we
can hardly rely on direct historical evidence for the simple reason that past transitions were rarely the
result of dedicated attempts (policy or otherwise) to realise them. What we can do is to try and combine
insights from two different bodies of research notably (1) the dynamics of historical transitions in which
various forms of policy intervention did play a role and (2) analysis of policy intervention to stimulate
innovation (if not transitions then incremental innovation). Above, we already briefly discussed the first
area and below we will deal with the second.
In policy science, three general governance paradigms are distinguished (see Table 1): (1) the
traditional top-down model with a central role for (national) government and hierarchical relations, (2) a
658 B. Elzen, A. Wieczorek / Technological Forecasting & Social Change 72 (2005) 651–661

Table 1
Different governance paradigms (based on [16])
Classic steering paradigm Market model Policy networks
(top-down, command-and-control) (bottom-up) (processes and networks)
Level of analysis Relationship between principal Relationship between Network of actors
and agent principal and local actors
Perspective Centralised, hierarchical Local actors Interactions between actors
organization
Characterisation of Hierarchical Autonomous Mutually dependent
relationships
Characterisation of Neutral implementation Self organization on the Interaction processes in which
interaction processes of formulated goals basis of autonomous information and resources are
decisions exchanged
Foundational scientific Classic political science Neo-classical economy Sociology, innovation studies,
disciplines (drational economic manT) neo-institutional political science
(dbounded rationalityT, uncertainty,
learning, interacting)
Governance instruments Formal rules, regulations Financial incentives Learning processes, network
and laws (subsidies, taxes) management, e.g. experiments,
demonstration projects, vision
building at scenario workshops
and foresight, network building
through seminars and strategic
conferences, public debates (. . .)

bottom-up or market model with a large degree of autonomy for local actors, and (3) a policy network
model of shared rule-making and agreements between interdependent actors with diverging values and
beliefs. These three governance paradigms not only differ in their basic philosophy, but also in their
instruments. Formal rules and regulations are common in the command-and-control paradigm, subsidies
and taxes in the market model, and network management, learning processes, experiments, and
interactive policy making in the third paradigm.
While the traditional model has been dominant at least since World War II, the market-model gained
prominence in the 1980s. In the 1990s interactive modes of policy making became more popular. In this
less hierarchical, more decentralised mode of governance, public and private actors interact in networks,
exchange information, and learn from each other.
With regard to innovation, the traditional model focused on changing the selection environment by
setting performance or safety standards. The market-model focused on stimulating variety, by giving
R&D subsidies or setting up technological research programs. Sometimes demand was stimulated by
providing buyer subsidies. The network paradigm positioned itself between supply-push and demand-
pull policies, by focusing on the learning processes and building of social networks necessary for
innovation. Metaphorically, while supply-push and demand-pull policies increase the pressure on
innovation, network and learning policies create the channels through which knowledge and products
flow.
The Jørgensen article [17] in this issue sheds some light on the importance of the policy
approaches described in the first two columns of Table 1. In the 1990s, a wave of liberalisation and
privatisation swept especially Europe and the US. Various utility functions like the supply of energy
 B. Elzen, A. Wieczorek / Technological Forecasting & Social Change 72 (2005) 651–661 659

and public transport were privatised because they were seen as inefficient and a more market-based
approach was expected to improve this. At the same time, these were sectors with large environmental
impacts and innovation was seen as a necessity to tackle the problems. Following a consistent faith in
the free market, top-down approaches to stimulate innovation were replaced by market-based
approaches. An interesting question then arises whether this actually supported a change towards
sustainability.
In his article, Jørgensen is very critical about it. The Danish government has had a long tradition of
using direct instruments to stimulate innovation on sustainable energy generation. As a result,
Denmark became one of the front-runners in the development of energy technologies based on
renewable sources, especially wind turbines. The privatisation and liberalisation wave, however, also
hit Denmark and, partly under the influence of the European Commission, the Danish government
turned to market-based instruments like a market for green energy and CO2 certificate trading.
Jørgensen analyses these different policy regimes from the perspective of their ability to support
technological innovation. His conclusion is that experience thus far provides little if any evidence that
simple, market-based models can facilitate the need for future energy technology innovations in a
satisfactory way. Thus, it seems, we need more complex and heterogeneous sets of measures.
The question then becomes: which sets of measures? Above, we proposed that interactive learning
and networks have to be crucial constituents of transition management or transition policy, which
suggests that the third, policy networks paradigm in the table above would be most appropriate. This
implies a need to engage various types of actors into the governance process and start a process of vision
building and learning. But how to do that?
In the transition management approach advocated by Kemp and Rotmans interactive learning is
considered crucial. To this end, a so-called transition arena should be organised that should engage in
a process of vision making [13]. van de Kerkhoff and Wieczorek [18] in this issue pose the question
of how this might be done, who should participate and how the debates should be organised. They
try to answer this question by using insights from the Dutch Climate OptiOns for the Long-term
(COOL) project that ran in the 1990s [19]. In this project a variety of stakeholders discussed the
issue of climate change, seeking to identify broadly supported possibilities via which the Netherlands
could tackle this problem. The authors come to the conclusion that this is far from easy as there are
usually conflicting requirements. There may be a conflict between seeking a representative
heterogeneity of stakeholders while on the other hand seeking homogeneity to keep some focus in
the discussions. There may also be conflict between inviting participants with a certain distance to
the issues at stake and people who are directly involved. Seeking to strike various balances in this
respect they propose a seven-step approach to explore and confront different visions and pathways
towards achieving these visions.
Their analysis demonstrates that an interactive approach may render fruitful exchanges of opinions in
which various actors come to acknowledge each other’s viewpoints but this does not necessarily mean
they will come to a consensus on longer-term visions and near term steps. Maybe consensus is even
more the exception than the rule. In cases of dissensus, individual actors will and will need to take their
own responsibility which, in the case of policy actors, may imply that they set some rules against the will
of (some) other actors.
Thus, although an interactive approach should be a crucial part of transition policy, it remains
necessary to also use the other two policy approaches listed in the table above as they all have a role to
play in the transition policy.
660 B. Elzen, A. Wieczorek / Technological Forecasting & Social Change 72 (2005) 651–661

The discussion above raises several important questions, including:

! What are the possibilities of new instruments from the policy networks paradigm for transitions?
Because this paradigm is rather new, its possibilities and effectiveness are far from clear.
! If all paradigms have a role to play in managing transitions, which is most appropriate under which
circumstances? Can this be related to different phases in transitions (e.g. invention and generation of
new options, the linking of novelties to existing regimes, the wider diffusion of the novelties and
transformation of the regime), which may require different policy interventions?
! What do the different policy paradigms imply for the role of different actors in supporting management
of transition especially the balance between dprivateT action and dpublicT intervention/support?

This last set of issues relates to a continuing debate over the roles of the public and private sectors in
supporting new technological developments and the associated changes in market and other societal
relations that assist the successful diffusion of new products and/or services. From the market-model
paradigm governments should only finance dpre-competitiveT research. Governments should not
interfere with the process of market development and diffusion, and leave that to private firms. In the
market-model it is unfashionable (compared to the 1960–1970s) for governments to dpick winnersT and
to be involved in all stages of the product/market development process.
From a network and learning paradigm, however, it might be worthwhile to revisit this public/private
divide. In the interest of more radical change required to realise sustainability, public intervention and
support might either be increased and/or take new forms. Since innovation in technologies and markets
cannot be limited to the actions of individual firms, but involves networks of firms and other actors
(especially users), the role of government agencies in dnetwork governanceT can be legitimately re-
examined. Possibly, public authorities need not to pick the winners themselves, but could take upon
them facilitating roles such as network builder, information exchanger, or agenda builder for the most
desirable directions to be followed (as in foresighting activities). In transition management there is
probably no clear-cut divide between the public and private spheres—the question is how to balance the
most effective levers of change.
The brief discussion above leads to the conclusion that there can be no such thing as a transition
policy. Transition policy (or transition management in the broader sense) consists of a variety of efforts
and actions, tuned to a specific situation and applied dynamically in the course of time as development
progresses. Transition policy is a matter of long breath, far longer than the typical cabinet period, which
in itself may already constitute a serious barrier to carry it out. This poses the challenge of developing
robust long-term policies that are relatively unsusceptible to whimsical political winds.4

References

[1] B. Elzen, F.W. Geels, K. Green (Eds.), System Innovation and the Transition to Sustainability: Theory, Evidence and
Policy, Edgar Elgar, Cheltenham, 2004.
[2] EU: European Transport Policy for 2010: Time to decide, white paper (2001).

4
Editor-in-chief’s note: In this connection the computer analysis of huge ensembles of scenarios by Lempert and Popper [20] at RAND,
reviewed in this Journal (issue 71/3 (March 2004)), may prove useful.
 B. Elzen, A. Wieczorek / Technological Forecasting & Social Change 72 (2005) 651–661 661

[3] G. Dosi, Technological paradigms and technological trajectories: a suggested interpretation of the determinants and
directions of technical change, Research Policy 6 (1982) 147 – 162.
[4] T.P. Hughes, The evolution of large technological systems, in: W.E. Bijker, T.P. Hughes, T. Pinch (Eds.), The Social
Construction of Technological Systems: New Directions in the Sociology and History of Technology, The MIT Press,
Cambridge, MA, 1987, pp. 51 – 82.
[5] R. Mayntz, T.P. Hughes (Eds.), The Development of Large Technical Systems, Campus Verlag, Frankfurt, and Westview
Press, Boulner, 1988.
[6] A. Rip, R. Kemp, Technological change, in: S. Rayner, E.L. Malone (Eds.), Human Choice and Climate Change, vol. 2,
Battelle Press, Columbus, OH, 1998, pp. 327 – 399.
[7] F.W. Geels, Technological Transitions an System Innovations: A Co-Evolutionary and Socio-Technical Analysis, Edgar
Elgar, Cheltenham, 2005.
[8] K. Green, C. Foster, Give peas a chance: transformations in food consumption and production systems, Technol. Forecast.
Soc. Change 72 (2005) 663–679.
[9] X. Olsthoorn, A. Wieczorek (Eds.), Sciences for Industrial Transformation: Views from different disciplines, Kluwer
Academic Publishers, Dordrecht, The Netherlands, 2005.
[10] F. Geels, Processes and patterns in transitions and system innovations: refining the co-evolutionary multi-level perspective,
Technol. Forecast. Soc. Change 72 (2005) 681–696.
[11] M.R. Taylor, E.S. Rubin, D.A. Hounshell, Control of SO2 emissions from power plants: a case of induced technological
innovation in the US, Technol. Forecast. Soc. Change 72 (2005) 697–718.
[12] J. Rotmans, R. Kemp, M. van Asselt, More evolution than revolution: transition management in public policy, Foresight 3
(1) (2001) 15 – 31.
[13] R. Kemp, J. Rotmans, Managing the transition to sustainable mobility, in: B. Elzen, F.W. Geels, K. Green (Eds.),
System Innovation and the Transition to Sustainability: Theory, Evidence and Policy, Edgar Elgar, Cheltenham, 2004,
pp. 137 – 167.
[14] G.S. Lynn, J.G. Morone, A.S. Paulson, Marketing and discontinuous innovation: the probe and learn process, California
Management Review 38 (3) (1996) 8 – 37.
[15] G.R. Teisman, J. Edelenbos, Getting through the dTwilight ZoneT: managing transitions through process-based, horizontal
and interactive governance, in: B. Elzen, F.W. Geels, K. Green (Eds.), System Innovation and the Transition to
Sustainability: Theory, Evidence and Policy, Edgar Elgar, Cheltenham, 2004, pp. 137 – 167.
[16] J.A. De Bruijn, W.J.M. Kickert, J.F.M. Koppenjan, Hoofdstuk 1: Inleiding: Beleidsnetwerken en overheidssturing, in:
J.F.M Koppenjan, J.A. De Bruijn, W.J.M. Kickert (Eds.), Netwerkmanagement In Het Openbaar Bestuur, Den Haag,
Vuga, 1993, pp. 11 – 30.
[17] U. Jørgensen, Energy sector in transition-technologies and regulatory policies in flux, Technol. Forecast. Soc. Change 72
(2005) 719–731.
[18] M. van de Kerkhof, A.J. Wieczorek, Learning and stakeholder participation in transition processes towards sustainability:
methodological considerations, Technol. Forecast. Soc. Change 72 (2005) 733–747.
[19] M. Hisschemöller, M. van de Kerkhof, M. Kok, R. Folkert, Climate OptiOns for the Long term (COOL): stakeholders’
views on 80% emission reduction, in: W. Vermeulen, A. Faaij, D. de Jager, M. Kok (Eds.), Global Warming and Social
Innovation: The Challenge of a Climate-Neutral Society, Earthscan, London, UK, 2002, pp. 196 – 216.
[20] R.J. Lempert, S.W. Popper, Shaping the Next One Hundred Years: New Methods for Quantitative, Long-Term Policy
Analysis, The RAND Corporation, Santa Monica, CA, 2003.

Boelie Elzen is a research associate at the STS Department of the University of Twente. His general research interest is in
understanding so-called transition processes and to explore ways to induce and stimulate transitions towards sustainability, also
called Transition Policy.

Anna Wieczorek is an executive officer of the Industrial Transformation project of the International Human Dimensions
Programme on Global Environmental Change (IHDP IT) hosted by the IVM. She fulfils a task of integrating and stimulating co-
operation among international and interdisciplinary scientists, whose work draws upon the dynamic interactions and mutual
dependencies between the producer (technological) and the consumer (market) domain, which have, for long, been studied
almost exclusively as rather isolated factors.