Ecological Economics 32 (2000) 319 – 332

www.elsevier.com/locate/ecolecon

ANALYSIS

Redefining innovation — eco-innovation research and the

contribution from ecological economics
Klaus Rennings *
Center for European Economic Research (ZEW), Research Area En6ironmental and Resource Economics,
En6ironmental Management, P.O. Box 103443, D-68034 Mannheim, Germany
Received 31 July 1998; received in revised form 29 July 1999; accepted 30 July 1999

Abstract

While innovation processes toward sustainable development (eco-innovations) have received increasing attention
during the past years, theoretical and methodological approaches to analyze these processes are poorly developed.
Against this background, the term eco-innovation is introduced in this paper addressing explicitly three kinds of
changes towards sustainable development: technological, social and institutional innovation. Secondly, the potential
contribution of neoclassical and (co-)evolutionary approaches from environmental and innovation economics to
eco-innovation research is discussed. Three peculiarities of eco-innovation are identified: the double externality
problem, the regulatory push/pull effect and the increasing importance of social and institutional innovation. While
the first two are widely ignored in innovation economics, the third is at the least not elaborated appropriately. The
consideration of these peculiarities may help to overcome market failure by establishing a specific eco-innovation
policy and to avoid a ‘technology bias’ through a broader understanding of innovation. Finally, perspectives for a
specific contribution of ecological economics to eco-innovation research are drawn. It is argued that methodological
pluralism as established in ecological economics would be very beneficial for eco-innovation research. A theoretical
framework integrating elements from both neoclassical and evolutionary approaches should be pursued in order to
consider the complexity of factors influencing innovation decisions as well as the specific role of regulatory
instruments. And the experience gathered in ecological economics integrating ecological, social and economic aspects
of sustainable development is highly useful for opening up innovation research to social and institutional changes.
© 2000 Elsevier Science B.V. All rights reserved.

Keywords: Innovation; Evolutionary economics; Technology; Economic incentives and disincentives

1. Introduction

Since the world community committed itself in

* Corresponding author. Tel.: +49-621-1235-207; fax: +
1992 in Rio to the principles of sustainable devel-
49-621-1235-226. opment, it has become more and more clear that
E-mail address: rennings@zew.de (K. Rennings) sustainability means long-term and far-reaching

0921-8009/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 9 2 1 - 8 0 0 9 ( 9 9 ) 0 0 1 1 2 - 3
320 K. Rennings / Ecological Economics 32 (2000) 319–332
changes in technologies, infrastructure, lifestyles
and institutions.
Thus, the importance of a better understanding
of innovation processes has several reasons:
“ The demand for drastic reductions of environ-
mental burdens, e.g. of greenhouse gases, im-
plies that adaptation within existing
technologies is not sufficient. Instead, regula-
tion strategies to effect ‘technology forcing’
and/or ‘technological regime shifts’ are needed.
“ Secondly, innovation is expected to offset bur-
dens and costs induced by environmental
regulations. Secondary benefits of an innova-
tion-friendly environmental policy are often
seen in reduced costs, increased competitive-
ness, creation of new markets for environmen-
tally desirable products and processes,
corresponding employment effects, etc. Al-
though these aspects have already been empha-
sized by Porter and van der Linde (1995a), the
Porter hypothesis postulating ‘innovation off-
sets’ of strict environmental policy is not em-
bedded in economic theory and is received with
scepticism among mainstream economists
(Jaffe and Palmer, 1996; Ulph, 1996).
“ New types of vehicles, renewable energy sys-
tems or corresponding infrastructure often
need at least a decade or more for invention,
for adaptation and for diffusion, respectively.
In total, it is realistic to assume time-scales of
half a century and more for major changes in
important economic and social sub-systems,
like technological regime shifts in energy and
transport systems. Thus, in situations far away
from the desired equilibrium, the importance of
analyzing transition and learning processes
moves into the foreground.
“ Moreover, many scenarios suppose that long-
term sustainability goals cannot be met by
progress in environmental technology and must
be supplemented by corresponding lifestyles,
e.g. through energy saving or changing mobil-
ity patterns, and institutional changes (ranging
from local networks to global organizations).
“ Inventing or adapting environmentally desir-
able processes or products is already part of
every day life for a large majority of firms and,
thus, a field of scientific research. As Cleff and
Rennings (1999a) have shown in a German
industry survey, about 80% of all innovating
firms are involved in environmental-friendly
innovation projects. It is hard to find even a
small or medium sized enterprise that has no
experience at all with substituting hazardous
substances, designing and using eco-efficient
products, saving energy, waste and material or
reducing emissions. Managing eco-innovation
is an increasingly important issue for many
firms.
“ Finally, innumerable sustainability programs
and initiatives have been set up to promote
innovative policy responses and corresponding
scientific research to improve the understand-
ing of global environmental change and it’s
relation to economic and social systems. Hav-
ing this in mind, together with long time-scales,
a careful valuation of experiences seems to be
crucial to identify key determinants and success
factors of innovation processes toward sustain-
ability, i.e. to analyze which experiments suc-
ceeded, which failed, why they failed and in
which phase.
Due to the fact that existing theoretical and
methodological frameworks do not address these
problems adequately, research needs can be iden-
tified to improve our understanding of innovation
processes toward sustainability in their different
dimensions, complex feedback mechanisms and
interrelations. Such a framework should be able
to give some guidelines about how to analyze the
driving forces of these processes in their different
characteristics and phases, to identify promising
examples as well as bad ones, and to give some
idea about their transferability to other contexts.
This paper intends to discuss the potential con-
tribution of neoclassical and (co-)evolutionary ap-
proaches from environmental and innovation
economics to fill this gap. A crucial question is
whether innovations toward sustainability can be
treated as normal innovations or if a specific
theory and policy are needed. The paper is orga-
nized as follows: Section 2 redefines the term
innovation considering its relation to sustainable
development. Section 3 describes main economic
approaches for analyzing innovation processes
and environmental policy, i.e. neoclassical and
 K. Rennings / Ecological Economics 32 (2000) 319–332
evolutionary concepts in environmental and inno-
vation economics. Finally, conclusions concerning
methodological pluralism in eco-innovation re-
search are drawn.
2. Redefining innovation
While ecological economists seem to be aware
of the need to redefine progress to meet the
challenge of sustainable development, no com-
parable effort has been made up to now to re-
define the term innovation. This is surprising
because innovation aspects play not only a major
role for national and international economic poli-
cies, but are also important elements of strategies
for sustainable development as described in the
introduction. Large budgets are spent for innova-
tion, particularly allocated to technology support
programs. From an ecological economics perspec-
tive, important questions are;
“ how to identify and promote technologies
which help to reach sustainability targets and
“ whether the focus on technologies is too nar-
row and a broader concept of innovation is
needed.
Thus, this section will redefine innovation with
regards to challenges of sustainable development.
2.1. Sustainable de6elopment
There is an ongoing debate on whether sustain-
able development can be defined operationally.
Some agree (overview in Rennings and Wiggering,
1997), others doubt or deny that it can (Nor-
gaard, 1994; Cary, 1998; Minsch 1997). Those
who doubt or deny understand sustainability
more as an heuristic idea, similar to ideas of
liberty and justice, guiding and orienting our
search rather than predicting its outcome.1
1 agriculture, forestry and households may be added. Several
As Cary (1998) (p.12) writes:
‘Sustainability is not a fixed ideal, but an evolutionary
process of improving the management of systems, through
improved understanding and knowledge. Analogous to
Darwin’s species evolution, the process is non-deterministic
with the end point not known in advance.’
321
However, with any of these interpretations of
sustainable development, at a certain point it is
necessary to give a more concrete idea about the
direction and problem areas of sustainability. In
its environmental report 1998, the German Coun-
cil of Environmental Advisers identified a consen-
sus on problem areas in seven major sustainability
concepts developed by European institutions2
(Table 1).
Obviously, these problem areas require progress
toward certain sustainability targets, which may
be different across regions, time-scales, target
groups, etc. The definition of problem areas and
the negotiation of targets can be analyzed and
evaluated by scientists, but decisions are made in
the political process. Ambitious goals as formu-
lated in the Toronto Resolution for greenhouse
gases may be watered down in the political pro-
cess and lead to modest agreements like those in
Table 1
Common areas of problems and sectors addressed in sustain-
ability conceptsa
Main problem areas Main sectors
Greenhouse effect Energy
Depletion ozone layer Mobility
Acidification Waste
Eutrophication
Toxic impacts on media/ecosystems
Toxic impacts on humans
Loss of biodiversity
Use of soil, land
Resource use
a
Source: SRU (1998), p. 88.
2
Six were from Germany (one NGO-report, three from the
German environmental protection agency, one from the gov-
ernment and one from a parliamentary commission) and one
from the European Commission. However, the problem areas
may be different in other contexts and countries. For example,
authors consider population policy as a further central element
of a policy of sustainability (e.g. Pestel and Radermacher,
1996). But compared with environmental problems this issue is
still not well addressed in most sustainability concepts and
strategies.
322 K. Rennings / Ecological Economics 32 (2000) 319–332
the Kyoto Protocol. However, in our context only
two features of a definition of sustainable devel-
opment are relevant: that it contains an ecologi-
cal, economic and social dimension and that even
modest sustainability targets, as fixed in the Ky-
oto Protocol, require substantial innovation.
2.2. Inno6ation
The conventional understanding of innovation
as defined in the Oslo-Manual of the OECD
(1997) (see also Hemmelskamp, 1997) distin-
guishes mainly between process, product and or-
ganizational innovation:
“ Process innovations occur when a given
amount of output (goods, services) can be pro-
duced with less input.
“ Product innovations require improvements to
existing goods (or services) or the development
of new goods. Product innovations in machin-
ery in one firm are often process innovations in
another firm.
“ Organizational innovations include new forms
of management, e.g. total quality management.
Innovation is different from invention, which is
an idea or a model for a new improved product or
process. In an economic sense, an invention be-
comes an innovation when the improved product
or process is first introduced to the market. The
third phase is the diffusion phase, when the inno-
vation is used and adopted over time.
For research on issues of sustainable develop-
ment, the OECD categories are useful but not
sufficient. It is useful because product and process
innovations include environmental technologies,
and because organizational innovations include
measures like eco-audits. Moreover, it is useful
that the OECD has recently extended innovation
research from industry to service sectors. This
means that eco-efficient services saving energy,
transport or waste are considered.
However, a weakness of the OECD definition is
that it does not explicitly distinguish environmen-
tal and non-environmental innovations. Conse-
quently, this distinction does not appear in
empirical innovation studies either. Finally, con-
sidering the challenges of sustainable develop-
ment, innovation in households and institutional
changes are missing in the OECD definition.
2.3. Inno6ation toward sustainable de6elopment
The general definition of innovation is neutral
concerning the content of change and open in all
directions. In contrast, putting emphasis on inno-
vation toward sustainable development is moti-
vated by concern about direction and content of
progress. Thus, the additional attribute of innova-
tions toward sustainability is that they reduce
environmental burdens at least in one item and,
thus, contribute to improving the situation in the
problem areas mentioned above. The interdisci-
plinary project ‘Innovation Impacts of Environ-
mental Policy Instruments’ has introduced the
term environmental innovation (short: eco-inno-
vation) and defined it very broadly as follows
(Klemmer et al., 1999):
Eco-innovations are all measures of relevant
actors (firms, politicians, unions, associations,
churches, private households) which;
“ develop new ideas, behavior, products and
processes, apply or introduce them and
“ which contribute to a reduction of environ-
mental burdens or to ecologically specified
sustainability targets.
Eco-innovations can be developed by firms or
non-profit organizations, they can be traded on
markets or not, their nature can be technological,
organizational, social or institutional. The follow-
ing paragraphs will look more specifically on the
distinction between;
“ technological,
“ organizational,
“ social and
“ institutional innovation.
Environmental technological measures can be
differentiated by those belonging to curative (e.g.
soil decontamination) or preventive environmen-
tal protection. As Fig. 1 illustrates, preventive
measures can be further subdivided into measures
of integrated and additive protection. The latter
 K. Rennings / Ecological Economics 32 (2000) 319–332
Fig. 1. Preventive environmental technologies. Source: Hohmeyer and Koschel (1995).
ones are also frequently referred to as end-of-
pipe-technology. Integrated environmental tech-
nology can be subdivided into product and
process integrated measures (Hemmelskamp,
1997). Organizational changes are, for example,
management instruments at the firm level like
eco-audits, which are of increasing importance for
innovation.
Changes of lifestyles and consumer behavior
are often defined as social innovations (Scherhorn
et al., 1997, p. 16). With regard to eco-innovation,
the term sustainable consumption patterns as
mentioned in the Rio Convention has received
increasing attention. Duchin (1999) argues that
the idea of social innovation is new
‘But one that is gathering an increasing fol-
lowing among social scientists and the public
more generally, that effective environmental
policy requires understanding not only techno-
logical but also lifestyle dynamics.’
323
Progress is still often understood simply as
innovation in firms, with a strong focus on tech-
nological progress. Due to the fact that many
problems of sustainable use of nature and land
are not primarily technological questions, this
may lead to a ‘technology bias’. Even more, Nor-
gaard (1994) (p. 16) identifies unsustainable devel-
opment itself as a result ‘from technology
outpacing changes in social organization’ and
postulates that, within a co-evolutionary
paradigm of a sustainable management of eco-
nomic and ecological systems, ‘incentives and reg-
ulations must evolve with technologies’. Natural
resources can often be characterized as open ac-
cess regimes, and unsustainable use stems from
inappropriate institutional arrangements. Innova-
tive institutional responses to problems of sustain-
ability may range from local networks and
agencies (e.g. for water resources of local rele-
vance) to new regimes of global governance (e.g.
an institution responsible for global climate and
324 K. Rennings / Ecological Economics 32 (2000) 319–332
biodiversity issues) and international trade (Ren-
nings et al., 1998; SRU, 1998, pp. 318 – 334).
Innovative institutions include improved decision
making through new ways of scientific assessment
and public participation. An example of an inno-
vative scientific network on the global level is the
Intergovernmental Panel on Climate Change
(IPCC), numerous other institutions for public
discourse upon environmental and technology im-
pact assessment have been established at the na-
tional, regional and local level. Thus, institutional
eco-innovations are often seen as a basic founda-
tion for a policy of sustainability (Freeman, 1992;
Minsch, 1997).
Examples for the different kinds of innovation
are: the 3-liter-car as an example of technical
eco-innovation, in contrast to higher-speed cars
with increased fuel consumption as an example of
non-environmental innovation. A shift of modal
split from car transport toward bicycling is an
example of social innovation, in contrast to a
modal shift towards air passenger traffic as an
example of non-environmental social innovation.
A network of NGOs, scientists, firms and public
authorities to promote sustainable transport or to
improve material flow management in a certain
region is an example of institutional eco-innova-
tion, while the same regional network co-ordinat-
ing an application for hosting Olympic games
would be a non-environmental innovation.
The distinctions between the different kinds of
innovations cannot be very sharp. Collective ac-
tions of households concerning sustainable con-
sumption patterns may be regarded as
institutional innovations, and the creation of en-
vironmental awareness in firms as social innova-
tion. Different kinds of innovation go
hand-in-hand or, using the terminology of Nor-
gaard, they co-evolve. In the words of Freeman
(1992) (p. 124):
‘Successful action depends on a combination
of advances in scientific understanding, appro-
priate political programs, social reforms and
other institutional changes, as well as on the
scale and direction of new investment. Organi-
zational and social innovations would always
have to accompany any technical innovations
and some would have to come first.’
3. Economic approaches to eco-innovation research
and policy
Having defined eco-innovation in a broad
sense, the question is how to analyze eco-innova-
tion and what ecological economics can con-
tribute to research. In this section, the most
relevant economic approaches for innovation re-
search, neoclassical and evolutionary concepts
and their possible contribution to eco-innovation
research will be discussed.
The distinction between neoclassical and evolu-
tionary approaches has to be justified. Using
neoclassical approaches mainly means that
methodological individualism is applied, which
can be incorporated into evolutionary approaches
too. Nevertheless, the distinction is useful for
analytical reasons because nearly all studies fol-
low only one approach and differ commonly from
the other by using different sets of assumptions.
It will be argued that ecological economics can
create an added value mainly by supporting inter-
disciplinary research. Methodological pluralism as
it is established in ecological economics would be
very valuable for eco-innovation research because
co-operation between the different disciplines (in-
novation and environmental sciences) and schools
(neoclassical vs. evolutionary concepts) is still
rather poor.
3.1. Eco-inno6ation in neoclassical economics
The issue of eco-innovation is placed at the
borderline between two different economic sub-
disciplines, environmental economics and innova-
tion economics. For an adequate analysis,
interdisciplinary research of both disciplines
would be very helpful. While environmental eco-
nomics tells how to assess environmental policy
instruments, innovation economics has led to in-
sights about the complexity of factors influencing
innovation decisions. Integrated approaches
should be able to identify and assess, in this
complex system, the role of state regulation to
stimulate innovation. However, such integrated
 K. Rennings / Ecological Economics 32 (2000) 319–332
approaches considering elements of both schools
are still hard to find. Thus, both approaches will
be described separately before synergies are
discussed.
3.1.1. En6ironmental economics
The superiority of market-based instruments
like taxes and tradable permits has been the basic
lesson from environmental economics concerning
innovation for a long time. These instruments
have been identified as environmental policy in-
struments with the highest dynamic efficiency (in-
novation efficiency). Their advantage is that they
give permanent incentives for further, cost-effi-
cient emissions reductions. By contrast, regulatory
regimes driven by technical standards (either in a
command-and-control system or in a regime of
voluntary agreements in which standards are ne-
gotiated between government and industry) are
not cost-efficient and the incentives for progress in
emission reduction vanish after the standards are
met.
However, several exceptions and modifications
to the rule have been made recently. For example,
the innovation efficiency of standards can be im-
proved substantially by ‘technology forcing’ in a
command-and-control regime (rules of permanent
reductions or long-term standards going beyond
existing technologies) and by repeated negotia-
tions in a regime of voluntary agreements (contin-
ued process of negotiations after each monitoring
phase) (Hohmeyer and Koschel, 1995; Brock-
mann, 1999). On the other hand, the innovation
efficiency of taxes may be watered down in the
political process. Total environmental costs for
industry are normally higher under a tax regime
than under alternative regimes of command-and-
control or negotiated agreements (because firms
have to pay for residual emissions and pollution).
This may lead to a tendency to impose relatively
low taxes with low innovation impacts. It is im-
portant to note that it is exactly the innovation-
friendly attribute of taxes (charging firms for
residual emissions), which may lead to this coun-
ter-effect (low tax level with low impacts) (Kemp
1997, p. 64). Further modifications to the rule
have been derived in general equilibrium models
of endogenous growth and in game theoretic
325
models. While the superiority of market-based
instruments has been confirmed for situations
with perfect competition and full information, the
situation changes under imperfect competition.
When firms gain ‘strategic advantages’ from inno-
vation, standards may be more appropriate for
stimulating innovation (Carraro, 1999).
In summary, the conclusion can be drawn that
the debate on dynamic efficiency is still open in
environmental economics. No instrument is gener-
ally preferable and the welfare gain of environ-
mental policy instruments depends on different
sets of circumstances (Fischer et al., 1998).
Against this background, Jaenicke (1999) criti-
cizes ‘instrumentalism’ in environmental policy,
i.e. the assumption that the choice of policy in-
struments determines the policy success. Accord-
ing to Jaenicke, specific instruments (taxes,
permits) are overestimated in the discussion while
important elements of successful environmental
policy are underestimated, as there are;
“ long-term goals and targets,
“ the mix of instruments,
“ different policy styles and
“ actor constellations.
Preliminary, it can be concluded that contribu-
tions on eco-innovation from environmental eco-
nomics suffer from a simple, mechanistic
stimulus-response model of regulation, neglecting
the complexity of determinants influencing inno-
vation decision in firms. These determinants will
be introduced in the following sections of this
paper.
3.1.2. Inno6ation economics
While in environmental economics methods
and strategies for the valuation and internaliza-
tion of negative external costs are developed, pos-
itive spillovers of basic R & D efforts in firms are
investigated in innovation economics. An impor-
tant peculiarity of eco-innovations is that they
produce positive spillovers in both the innovation
and diffusion phase. Positive spillovers in the
diffusion phase appear due to a smaller amount
of external costs compared to competing goods
and services on the market. In the following, this
peculiarity of eco-innovations will be called the
double externality problem.
326 K. Rennings / Ecological Economics 32 (2000) 319–332
Fig. 2. Determinants of eco-innovations. *OSH=Occupational Safety and Health.
The double externality problem reduces the in-
centives for firms to invest in eco-innovations.
Therefore environmental policy and innovation
policy should be co-ordinated. Innovation policy
can help to cut the costs of technological, institu-
tional and social innovation, especially in the
phases of invention and market introduction, e.g.
by financial support for pilot projects, and in the
diffusion phase it may help to improve the perfor-
mance characteristics of eco-innovations. At least
in the diffusion phase, however, environmental
policy is responsible for internalizing external
costs imposed by competing, non-ecological prod-
ucts or services. As long as markets do not punish
environmental harmful impacts, competition be-
tween environmental and non-environmental in-
novation is distorted.
Due to the fact that both externalities result in
a sub-optimal investment in eco-innovations, the
double externality problem induces a second pe-
culiarity: the importance of the regulatory frame-
work as a key determinant for eco-innovative
behavior in firms, households and other institu-
tions. The main discussion in innovation econom-
ics has been whether technological innovation has
been driven by technological development (tech-
nology push) or by demand factors (market pull).
Empirical evidence has shown that both are rele-
vant (Pavitt 1984). With regard to eco-innovation,
new eco-efficient technologies can be subsumed
under technology push factors, while preferences
for environmentally friendly products or image
can be subsumed under market pull factors. Due
to the externality problem of eco-innovations, the
traditional discussion of innovation economists
has to be extended to the influence of the regula-
tory framework. In the following, this peculiarity
of eco-innovations will be called the regulatory
push/pull-effect. Fig. 2 illustrates the determinants
of eco-innovation. As empirical evidence shows
(Green et al., 1994; Porter and van der Linde,
1995a,b; Kemp, 1997; Faucheux and Nicolai
1998), the regulatory framework and especially
environmental policy have a strong impact on
eco-innovation. Eco-innovations are, in contrast
to such technologies as microelectronics and
telecommunications, normally not self-enforcing.
Because factors of technology push and market
pull alone do not seem to be strong enough,
eco-innovations need specific regulatory support.
Cleff and Rennings (1999a,b) have identified
eco-innovators using data from the German Inno-
vation Panel and analyzed the different factors
influencing eco-innovation decisions in firms. The
 K. Rennings / Ecological Economics 32 (2000) 319–332 327

study shows that, within their innovation goals, Their contribution to a theory and policy of eco-
eco-innovative firms attach significantly higher innovation will be discussed in the next section.
importance to the goals of cost reduction and
3.2. Eco-inno6ation in (co-)e6olutionary
total quality management (TQM) than other in-
approaches
novators. This gives evidence to the Porter hy-
pothesis that firms understand eco-efficiency as a
While deterministic neoclassical models have
part of total efficiency. With regard to integrated
their merits, especially for analyzing marginal or
technologies, Cleff and Rennings differ between
incremental changes induced by different kinds of
ecological product- and process-innovations. En-
incentives, they are of limited value for the analy-
vironmental product innovation is significantly
sis of more radical changes of technological sys-
driven by the strategic market behavior of firms
tems including the organizational and societal
(market pull effect), while environmental process-
context. According to Freeman (1992) (pp. 77–
innovation is more driven by regulation (regula-
81), incremental innovations can be characterized
tory push/pull effect). As for the influence of
as continuous improvements of existing techno-
individual policy instruments on environmental
logical systems (i.e. they fit in existing input–out-
innovation, influence from ‘soft’ regulatory instru-
put tables) while radical innovations are
ments (environmental liability, eco-audits, volun-
discontinuous (i.e. they require new lines and
tary commitments) and from eco-labels can be
columns in input–output tables).
found. These instruments enable firms to use their
Evolutionary approaches have therefore been
environmental performance in their marketing
developed to open up the ‘black box’ of surprises
strategies or in negotiations with the state. Envi-
being connected with radical changes: unpre-
ronmentally innovative firms seem to be less de-
dictable interactions of sub-systems, irreversibil-
pendent on ‘hard’ state regulation than other,
ity, path-dependency, lock-in effects of
more passive firms. Thus, ‘soft’ and voluntary
technological trajectories or bifurcation. Evolu-
environmental policy measures may be sufficient
tionary approaches are more interested in the
for pioneers. However, ‘hard’ measures (com-
analysis of transition and learning processes than
mand and control instruments, duties) seem to be
in equilibrium states and assume bounded ratio-
still necessary for a diffusion of integrated mea-
nality and rules of thumb rather than optimiza-
sures to non-innovative firms.
tion. The main methods are case studies and ex
Using industry micro-data either for technolog-
post analysis because predictions regarding which
ical or organizational innovations, such studies
option will succeed are recognized as being
may improve our understanding of firm decisions
impossible.
concerning short-term, incremental changes. They
should be supplemented by additional surveys on 3.2.1. Variation, selection and co-e6olution
eco-innovation in the service sector. With regard The biological terms of selection and variation
to strategies of de-materialization or de-car- are used to describe the innovation process. In-
bonization of consumption and production pat- ventions are variations which succeed or fail in
terns within a policy of sustainability, innovation the evolutionary process due to selection criteria
in the service sector plays a key role. Surveys of their environment. According to Freeman
should be supplemented by case studies analyzing (1992) (pp. 123–127), the selection environment
the success and failure of interrelated technologi- of the innovation process can be divided into
cal, social and institutional eco-innovation. three categories:
For long-term innovation processes including “ Natural environment. Man-made environmen-
more radical changes, however, neoclassical mod- tal problems or external forces may put selec-
els assuming marginal changes and equilibrium tive pressure on society to create new
situations may be too narrow. Broader evolution- technologies, e.g. phasing out CFCs to protect
ary approaches have been developed to improve the ozone layer or forcing energy saving tech-
our understanding of radical system changes. nologies to mitigate climate change.
328 K. Rennings / Ecological Economics 32 (2000) 319–332
“ Built environment. The built environment con-
sists of physical assets, i.e. the existing in-
frastructure. The built environment needs
decades to depreciate, thus, slowing down in-
novation and diffusion processes.
“ Institutional environment. Profitability can be
identified as a key selection criterion in market
economies.
It should be noted that the variation-selection-
terminology focuses only on technological innova-
tions and does not explicitly consider complex
feedback mechanisms between variations and the
selection environment. This seems to be inade-
quate bearing in mind that an innovation is not
only selected by the environment but also changes
the environment by selective pressures. CFCs
have depleted the ozone layer and led to the
invention of CFC-substitutes, changes of institu-
tions and consumer behavior. These responses
(such as ‘soft’ CFCs) put again selective pressures
on the environment, raising concern for improved
eco-innovations. Similar feedback mechanisms
can be observed in energy policy where eco-inno-
vations such as cleaner fossils, safer nuclear, ratio-
nal use of energy and renewables are responses to
environmental pressures (scarcity of fossil fuels,
air pollution, greenhouse effect, nuclear risks) but
also change the environment by selective
pressures.
In ecological economics, these complex feed-
back mechanisms have been addressed by the
co-evolutionary paradigm as defined by Norgaard
(1984) (p. 161):
‘In biology, co-evolution refers to an evolu-
tionary process based on reciprocal responses
between two interacting species. … The concept
can be broadened to encompass any ongoing
feedback process between two evolving systems,
including social and ecological systems. … So-
ciosystems and ecosystems are maintained
through numerous feedback mechanisms. Co-
evolution occurs when at least one feedback is
changed, which then initiates a reciprocal pro-
cess of change.’
Having in mind the danger of a technology bias
as mentioned in Section 2, the co-evolutionary
framework seems to be more appropriate to ana-
lyze eco-innovations for at least two reasons;
“ it includes all sub-systems, i.e. co-evolving so-
cial, ecological and institutional systems avoid-
ing any ranking of their importance, and
“ it underscores the importance of their
interactions.
The history of pesticide policy and the phasing
out of CFCs are textbook examples of co-evolu-
tionary innovation processes highlighting the im-
portance of interactions between technological,
social and institutional innovations. In the CFC
phase-out, the Montreal Protocol has been a key
success factor illustrating the importance of insti-
tutional innovations (see for details on pesticide
policy Norgaard, 1994, pp. 23–28; on the CFC
phase-out in Germany and the United States
Kuehn and Osório-Peters, 1999).
Freeman’s emphasis on the crucial role of insti-
tutional and social re-organization (cited in Sec-
tion 2) within a paradigm of ‘green’ innovation
shows that he is well aware of the need for a
broader approach. However, the co-evolutionary
approach has not yet been elaborated for specific
purposes of eco-innovation research.3 Thus, a re-
search need can be identified for opening up
evolutionary approaches in innovation economics
to co-evolving ecological, institutional and tech-
nological systems.
3.2.2. Technological change
As already mentioned, evolutionary approaches
have focused on technological innovation in the
past. Nevertheless, these studies have given valu-
able insights about determinants of technological
change, which are relevant for eco-innovation
research.
Due to the pressures of the selection environ-
ment a certain technology may become a domi-
nant ‘technological paradigm’. Advantages in
transaction costs, learning curves, economies of
scale, superior cost-benefit-ratios and a good fit
3
Although Kemp (1997) (p. 3 and 276) uses the term
co-evolution, he does not explain or elaborate this terminology
and concentrates on technological innovations.
 K. Rennings / Ecological Economics 32 (2000) 319–332
with existing lifestyles, technologies, infrastruc-
tures or networks result in path-dependencies or
technological trajectories (Dosi, 1988), i.e. to
lock-in effects of a technology excluding other
evolutionary options. Examples of technological
paradigms are oil-based chemistry and semi-
conductors.
Kemp (1997) (pp. 279 – 289) describes determi-
nants and success factors of technological change
as follows:
“ Determinants are new scientific insights which
open up new technological opportunities,
pressing technological needs (e.g. technological
bottlenecks to further emission reductions
through incremental improvements of end-of-
pipe technologies, high costs of further ad-
vances within a technical design, such as
carbon dioxide reductions within fossil energy
technologies, changes in demand, scarcity of
materials, or labor conflict) and entrepreneurial
activities and institutional support for radically
original technologies.
“ Important success factors of radical technologi-
cal change are early market niches and the use
of available knowledge and techniques, i.e. a
certain compatibility with existing know how,
experience and infrastructure.
Having identified these success factors, Kemp
(1997) (p. 310) suggests fostering technological
change by a policy of strategic niche management,
i.e. the ‘creation of protected spaces for promising
technologies that we want to point out’. The idea
is to install temporary pilot markets protected by
subsidies or other regulatory measures. Examples
are:
“ The success stories of wind energy markets in
Denmark and Germany which are especially
interesting for a close cooperation between en-
vironmental and eco-innovation policy, i.e. be-
tween economic incentives (subsidies in
Germany, energy tax in Denmark) and tech-
nology support programs (Hemmelskamp,
1999).
“ The ‘Los Angeles Initiative’ requiring that
zero-emission cars must account for 2 – 10% of
new car production in the 1998 – 2003 period as
creating a temporary protected area for electric
vehicles (Templin, 1991, p. 310).
329
“ The idea of a ‘Renewables Portfolio Standard’
under which every retail power supplier would
be required to purchase renewable energy cred-
its equivalent to some percentage of its total
energy sales (Rader and Norgaard, 1996).
The examples show that a policy fostering tech-
nological eco-innovations cannot be reduced to
technological support programs nor to conven-
tional environmental policy measures, but has to
find intelligent combinations of both. The prob-
lem is, of course, to find a balance between pro-
tection and selection pressure. However, some
protection may be necessary, even in the diffusion
phase, due to the degree of existing external costs
not yet internalized by environmental policy.
Thus, close co-ordination between environmental
policy and eco-innovation policy will be
necessary.
3.2.3. Perspecti6es
Evolutionary approaches seem to be very useful
for providing additional insight into radical tech-
nological change. Compared to neoclassical eco-
nomics they follow a broader approach as they
allow surprises and consider technological path-
dependencies. It would be worthwhile, however,
to open up the evolutionary framework to ecolog-
ical irreversibility and to strengthen social and
institutional innovations. This would be an im-
portant contribution from ecological economics
to eco-innovation research, avoiding a ‘technol-
ogy bias’ and unsustainable development.
Moreover, it seems to be beneficial to link
neoclassical and evolutionary models in further
research. For example, Kemp (1997) has com-
bined neoclassical and evolutionary approaches of
eco-innovation to a certain extent. He introduced
uncertainty, specific technology characteristics
and shifting consumer preferences into neoclassi-
cal models and rational choice and optimization
into the evolutionary discussion of technological
regime shifts.
4. Perspectives for methodological pluralism in eco-
innovation research
Both neoclassical and (co-)evolutionary ap-
proaches have their merits and limits concerning a
330 K. Rennings / Ecological Economics 32 (2000) 319–332
theory and policy of eco-innovation. The evolu-
tionary approach is broader and tries to take a
real-life picture of a certain innovation process
avoiding any generalization. This is an important
advantage as far as stochastic phenomena have to
be understood, which are common especially for
radical innovation. Thus, evolutionary ap-
proaches are appropriate as far as long-term,
radical changes including path-dependencies, irre-
versibility, transition processes, discontinuous and
unpredictable events are concerned.
Recently, international case studies on the im-
pact of environmental policy on eco-innovation
have been carried out with an approach that
combines elements of both evolutionary econom-
ics and policy analysis. Based on Jaenicke’s criti-
cism of ‘instrumentalism’ in environmental policy
(see Section 3.1.1) and on seven international case
studies, Blazejczak et al. (1999) have identified
three essential success factors for innovation-ori-
ented environmental policy: instruments, policy
styles and actor constellation. This ‘policy style’-
approach postulates that concerning instruments,
important success factors are a mix of instru-
ments, strong economic incentives, strategic plan-
ning and a consideration of the different phases of
the innovation process. With regard to policy
style, calculable goals (e.g. long-term national en-
vironmental policy plans) are an important ele-
ment to stimulate innovation especially when they
are combined with flexible means. Actor constel-
lations should include inter- and intra-policy inte-
gration, co-ordination between regulators and
target group and between stakeholders along the
production chain.
However, not every innovation process can be
assumed to be random or unpredictable, espe-
cially incremental innovation, which is largely de-
termined by a given technological trajectory. If
the behavior of actors or sub-systems and rela-
tions between them can be assumed to be rational
and stable, methods derived from neoclassical the-
ory are preferable for eco-innovation research.
They offer quantitative tools and are most elabo-
rated to analyze the efficiency of incentive sys-
tems, which is essential for stimulating
innovation. Furthermore, neoclassical approaches
explain two important peculiarities of eco-innova-
tion: the double externality problem and the regu-
latory push/pull effect.
The co-evolutionary approach has been used to
explain the third peculiarity of eco-innovation
emphasizing the interactions of ecological, social
and institutional systems. Both the evolutionary
and the neoclassical framework should be opened
up to interactions with ecological systems (e.g. to
consider not only the impact of regulation on
eco-innovation but also the ecological impact of
eco-innovation) and to strengthen the importance
of social and institutional innovations.
Some may wonder if it makes sense to redefine
every kind of technical, institutional and socioeco-
nomic change or reform as innovation. However,
in our context, the only relevant criteria for valu-
ing change is that it is somehow new and a certain
likelihood that it leads in the desired direction. A
more restrictive selection and support of options,
e.g. a focus on technologies, is explicitly not in-
tended as long as a diversity of other options
exists which might be supplementary, superior,
etc. Thus, innovation policy should open up a
narrow technological definition of innovation to
all kinds of organizational, behavioral and institu-
tional change.
A look on the ongoing German innovation
support program, ‘Demonstration Projects for
Sustainable Development’ (BMBF/GSF 1999)
may illustrate the existing theoretical vacuum in
the field of social and institutional innovation.
Within this program, the German government
supports 14 demonstration projects in the area of
social and institutional eco-innovation, i.e.:
“ Regional projects on sustainable agriculture.
The ministry supports networks aiming at the
regional distribution of ecological food.
“ Regional projects on material flow manage-
ment. The ministry supports networks of, e.g.
firms optimizing their material flow manage-
ment by exchanging materials and waste.
“ Projects concerning specific regional needs, e.g.
optimization of material flow management in a
regional construction sector.
While such research programs and demonstra-
tion projects on social and institutional innova-
tion are already going on, theoretical and
methodological approaches for their evaluation
 K. Rennings / Ecological Economics 32 (2000) 319–332
are missing. Thus, it seems to be important to
elaborate approaches like strategic niche manage-
ment for applications to social and insitutional
eco-innovation. Due to the fact that eco-innova-
tion policy requires the implementation of effi-
cient incentive systems and coordination between
research and environmental policy, neoclassical
approaches including new institutional economics
(game theory, public choice) are important, too.
However, the theoretical and empirical work on
eco-innovation is still in its beginning. Theoretical
approaches and methods are required as well as
data. Essential material for further research would
be data-banks on the eco-innovation behavior of
firms, households and stakeholders in networks.
Such data-banks have to be developed by regular
surveys. Other issues for a research agenda are
ecological, social and economic impacts of eco-in-
novations (innovation impact assessment) as for
example impacts on employment, eco-innovation
in the service sector, or national and regional
eco-innovation systems.
Some elements of eco-innovation theory and
policy have not yet been mentioned or elaborated
in this paper. They include management ap-
proaches in business administration (management
of eco-innovation as outlined, e.g. by Fuzzler,
1996). However, within a broader co-evolutionary
paradigm, conceptual and methodological plural-
ism and interdisciplinary research are welcome.
Approaches from business administration and
other disciplines enrich the discussion. Synergies,
conflicts and complementarity between the con-
cepts may be an issue for further research.
Acknowledgements
A former version of this paper (title: ‘Towards
a theory and policy of eco-innovation — neoclas-
sical and (co-)evolutionary perspectives’) was
written during a visiting scholarship at the Uni-
versity of California in Berkeley in the summer of
1998. Work has been made possible by internal
funding from ZEW. However, the work is closely
related to a sub-task of the project ‘Innovation
impacts of environmental policy instruments’
(German acronym: FIU) commissioned by the
331
German Ministry of Education and Research
(BMBF). I am grateful to Jens Hemmelskamp
(Joint Research Center, Sevilla), Richard Nor-
gaard (University of California, Berkeley), my
colleagues Karl Ludwig Brockmann and Henrike
Koschel, Ernst Mohr and two anonymous review-
ers for stimulating discussions and helpful
comments.
References
BMBF, GSF (Editors), 1999. Innovative approaches for re-
gional economic development — demonstration projects
for sustainable development 1998 – 2002. Federal Ministry
of Research and Education, Research Center for Environ-
ment and Health, Bonn, Munich, http://www.gsf.de/
PTUKF/deutsch.html.
Blazejczak, J., Edler, D., Hemmelskamp, J., Jaenicke, M.,
1999. Umweltpolitik und Innovation – Politikmuster und
Innovationswirkungen im internationalen Vergleich.
Zeitschrift für Umweltpolitik und Umweltrecht (ZfU), 1/
99, pp. 1 – 32.
Brockmann, K.L., 1999. Anreizmechanismen und Innovation-
swirkungen ‘freiwilliger Selbstverpflichtungen’ im
Umweltschutz-eine spieltheoretische Analyze. In: Ren-
nings, K. (Ed.), Innovation durch Umweltpolitik. ZEW
Wirtschaftsanalysen. Nomos Verlag, Baden Baden, pp.
103 – 146.
Carraro, C., 1999. Environmental technological innovation
and diffusion: model analysis. In: Hemmelskamp, J.,
Leone, F., Rennings, K. (Eds.), Innovation-oriented Envi-
ronmental Regulation: Theoretical Approaches and Empir-
ical Analysis. Physica Verlag, Heidelberg, New York.
Cary, J. 1998. Institutional innovation in natural resource
management in Australia: the triumph of creativity over
adversity. In: Abstracts of the Conference ‘Knowledge
Generation and transfer: Implications for Agriculture in
the 21st Century’. University of California-Berkeley, June
18 – 19, pp. 11 – 13.
Cleff, T., Rennings, K., 1999a. Determinants of environmental
innovation-empirical evidence from the Mannheim Innova-
tion Panel and an additional telephone survey. In:
Hemmelskamp, J., Leone, F., Rennings, K. (Eds.), Innova-
tion-oriented Environmental Regulation: Theoretical Ap-
proaches and Empirical Analysis. Physica Verlag,
Heidelberg, New York.
Cleff, T. and Rennings, K., 1999b. Determinants of environ-
mental process and product innovation — evidence from
the Mannheim Innovation Panel and a follow-up telephone
survey, Special issue on integrated product policy, Karl,
M., Orwat, C. (Eds), European Environment 9 (5), in
press.
Dosi, G., 1988. The nature of the innovation process. In: Dosi,
G., Freeman, C., Nelson, R., Silverberg, G., Soete, L.
(Eds.), Technical Change and Economic Theory. Printer
Publishers, London, pp. 221 – 238.
332 K. Rennings / Ecological Economics 32 (2000) 319–332
Duchin, F., 1999. Reducing pressures on the environment:
forward-looking solutions and the role of formal models.
In: Hemmelskamp, J., Leone, F., Rennings, K. (Eds.),
Innovation-oriented Environmental Regulation: Theoreti-
cal Approaches and Empirical Analysis. Physica Verlag,
Heidelberg, New York.
Faucheux, S., Nicolai, I., 1998. Environmental technological
change and governance in sustainable development policy.
Ecol. Econom. 27 (1998), 243–256.
Fischer, C., Parry, I.W.H., Pizer, W.H., 1998. Instrument
choice for environmental protection when technological
innovation is endogenous. Resources for the Future, Dis-
cussion Paper 99-04, Washington.
Freeman, C., 1992. The Economics of Hope. Pinter Publishers,
London, New York.
Fuzzler, C., 1996. Driving Eco-Innovation. Pitman Publishing,
London.
Green, K., McMeekin, A., Irwin, A., 1994. Technological
trajectories and R & D for environmental innovation in
UK firms. Futures 26, 1047–1059.
Hemmelskamp, J., 1997. Environmental Policy Instruments
and their Effects on Innovation. Eur. Plan. Stud. 2, 177–
194.
Hemmelskamp, J., 1999. Innovationswirkungen der Umwelt-
politik im Windenergiebereich. In: Klemmer, P. (Ed.), In-
novationen und Umwelt. Analytica-Verlag, Berlin, pp.
81– 112.
Hohmeyer, O., and Koschel, H., 1995. Umweltpolitische In-
strumente zur Förderung des Einsatzes integrierter
Umwelttechnik. Final Report of a Study of ZEW commis-
sioned by the Büro für Technikfolgenabschätzung beim
Deutschen Bundestag (TAB). Mannheim, copied
manuscript.
Jaenicke, M., 1999. Policy patterns and innovation impacts: an
international comparison. Paper presented at the Interna-
tional Conference ‘Innovation-oriented Environmental
Regulation – Theoretical Approaches and Empirical Anal-
ysis’. Potsdam, 27 – 29 May 1999, organized by IPTS,
Spain, ZEW, Germany and MERIT, Netherlands.
Jaffe, A.B., Palmer, K., 1996. Environmental Regulation and
Innovation: A Panel Data Study. National Bureau of
Economic Research. Working Paper 5545, Washington
D.C.
Kemp, R., 1997. Environmental Policy and Technical Change.
Edward Elgar, Cheltenham, Brookfield.
Klemmer, P., Lehr, U. and Löbbe, K., 1999. Environmental
Innovation. Volume 3 of publications from a Joint Project
on Innovation Impacts of Environmental Policy Instru-
ments. Synthesis Report of a project commissioned by the
German Ministry of Research and Technology (BMBF),
Analytica-Verlag, Berlin.
Kuehn, I., Osório-Peters, S., 1999. Innovationswirkungen frei-
williger Selbstverpflichtungen imUmweltschutz – spieltheo-
retische Analysen und Fallstudien zum Klimaschutz in
Deutschland un den Niederlanden. In: Klemmer, P. (Ed.),
Innovationen und Umwelt. Analytica-Verlag, Berlin, pp.
329 – 344.
Minsch, J., 1997. Nachhaltigkeit und institutionelle Innova-
tionen. In: Rennnings, K., Hohmeyer, O. (Eds.), Nach-
haltigkeit. Nomos-Verlag, Baden-Baden, pp. 297 – 329.
Norgaard, R., 1984. Co-evolutionary development potential.
Land Econom. 60 (2), 160 – 173.
Norgaard, R., 1994. Development Betrayed. Routledge, Lon-
don and New York.
Pavitt, K., 1984. Sectoral patterns of technical change: towards
an taxonomy and a theory. Res. Pol. 13, 343 – 373.
Pestel, R., Radermacher, F.-J., 1996. Robuste Pfade zur
globalen Stabilitaet. In: Morath, K. (Ed.), Welt im Wan-
del-Wege zu dauerhaft-umweltgerechtem Wirtschaften.
Frankfurter Institut, Bad Homburg.
OECD, 1997. OECD Proposed Guidelines for Collecting and
Interpreting Technological Innovation Data – Oslo-Man-
ual, OECD/Eurostat, Paris.
Porter, M.E., van der Linde, C., 1995a. Towards a new
conception of the environment-competitiveness relation-
ship. J. Econom. Perspect. 9 (4), 97 – 118.
Porter, M.E., van der Linde C., 1995b. Green and Competi-
tive. Ending the Stalemate. Harvard Business Review, Sep-
tember – October 1995, pp. 120 – 134.
Rader, N. and Norgaard, R., 1996. Efficiency and Sustainabil-
ity in Restructured Electricity Markets: The Renewables
Portfolio Standard. Electricity Journal, July 1996 9 (6) pp.
37 – 49.
Rennings, K., Koschel, H., Brockmann, K.L. and Kuehn, I.,
1998. A Regulatory Framework for a Policy of Sustain-
ability-Lessons from the Neoliberal School. Ecological
Economics, 28 (1999), pp. 197 – 212.
Rennings, K., Wiggering, H., 1997. Steps towards indicators
of sustainable development — linking economic and eco-
logical concepts. Ecol. Econom. 20, 25 – 36.
Scherhorn, G., Reisch, L., Schroedl, S., 1997. Wege zu nach-
haltigen Konsummustern-Ueberblick ueber den Stand der
Forschung und vorrangige Forschungsthemen. Metropolis-
Verlag, Marburg.
SRU-Rat von Sachverstaendigen fuer Umweltfragen (German
Council of Environmenta Advisers), 1998. Umweltgutach-
ten 1998. Verlag Metzler-Poeschel, Stuttgart.
Templin, N., 1991. California Rules Push Car Makers to
Clean Up Act. Wall Street Journal Europe, March 27,
1991.
Ulph, A.M., 1996. Strategic environmental policy and interna-
tional competitiveness. In: Siebert, H. (Ed.), Elemente einer
rationalen Umweltpolitik. Thesen fuer eine umweltpolitis-
che Neuorientierung. Mohr Verlag, Tuebingen, pp. 337 –
376.