Environmental macro-indicators of innovation

Environmental macro-indicators of innovation: policy, assessment and monitoring - EMInInn Final Report Lead authors: •

Will McDowall (UCL)

•

Fernando J. Diaz Lopez (TNO)

•

Laura Seiffert (TNO)

With contributions from: • • • • • • • •

Rene Kemp (MERIT) Serdar Turkeli (MERIT) Roberto Zoboli (CERIS) Paul Ekins (UCL) Olga Ivanova (TNO) Mohammed Chaim (TNO) David Font Vivanco (CML) Ruben Huele (CML)

Environmental Macro Indicators of Innovation EMInInn is a collaborative project funded by the EU’s Seventh Framework Program – Theme ENV.2011.3.1.9-3 Grant agreement no: 283002 Deliverable number Revision number Date of current draft Due date of deliverable Actual submission date Dissemination level

D10.4 1 08.04.2015

PU

Environmental Macro Indicators of Innovation

Title


Projekt ID

283002

Call

Start Date

1.11.2011

Duration

FP7-ENV-2011-ECOINNOVATION 42 months O S

Programme

FP7

Rdg

RTD

Abstract

EMInInn aims at assessing the environmental impacts associated with innovation. In a first step EMInInn will assemble and set out coherently, on the one hand, macro-indicators and data of environmental impacts and, on the other hand, indicators and data to measure innovations. The definitions and delineations will be the basis for selecting appropriate analytical frameworks to operationalize assessments of environmental impacts associated with innovation on a macro scale. EMInInn will incorporate and integrate a number of advanced analytical approaches for the ex post assessment of the macro-environmental impacts of innovation. This methodology will be applied in different areas of technological innovation: • Energy sources and conversion technologies • Information and Communication Technologies • Transport • Built environment and buildings • Waste management EMInInn aims at developing an analytical framework for assessing environmental impacts of established as well as emerging technologies. In selected cases options for scenarios to model burden-shifting and rebound effects will be explored. EMInInn will strengthen the science-policy link. An advisory board with experts from different governance levels will be to advise the EMInInn researchers how to link the research and dissemination with ongoing and upcoming policy initiatives and research. A number of workshops and publications will allow interaction with experts, stakeholders and policy-makers. In that context EMInInn will address EU policies, which affect three major fields of environmental impact: • resources and waste, • energy and climate, as well as • land-use and biodiversity. By improving environmental assessments of innovation as well as policy-oriented interactions and outputs, EMInInn will generate contributions for improving EU-policies for a transition towards a more sustainable Europe and thus contribute to the flagship initiatives for a Resource Efficient Europe and the Innovation Union. Participants

• • • • • • •

Wuppertal Institute for Climate, Environment and Energy (WI) - Coordinator UCL Energy Institute, University College London (UCL) Institute of Environmental Sciences, Leiden University (LU-CML) Netherlands Organization for Applied Scientific Research (TNO) Institute for Economic Research on Firms and Growth (CERIS) Swedish Environmental Research Institute (IVL) Maastricht Economic and Social Research and Training Centre on Innovation and Technology, Maastricht University (UM-MERIT) For more information please visit the project website: http://www.emininn.eu

2


Contents 1

Executive Summary ......................................................................................................................... 4

2

Introduction .................................................................................................................................... 8

3

Concepts and analytical model to understand the environmental impacts of eco-innovation ... 13

4

Current policy practice .................................................................................................................. 43

5

Identifying the environmental pressures from specific innovations ............................................ 65

6 Monitoring eco-innovation at the macro-level: applications and methodological advances in EMInInn ................................................................................................................................................. 78 7 When does the innovation process drive reductions in environmental pressures? Drivers, barriers and framework conditions of eco-innovation ......................................................................... 84 8

Summary of insights from case studies and from macro-modelling .......................................... 109

9

Policy implications and main messages ...................................................................................... 114

10

Annex on results of macro-modelling: rebound effect and sensitivity testing ...................... 119

11

Full list of EMInInn Deliverables and reports .......................................................................... 142

12

References .............................................................................................................................. 144

3


1 Executive Summary Eco-innovation holds the promise of facilitating the achievement of environmental goals while supporting economic welfare. This idea has been appealing to policy makers and politicians, as it resolves an apparent trade-off between environmental quality and economic prosperity. Eco-innovation policies aim to stimulate the development and deployment of innovations that reduce environmental pressures compared to a relevant alternative. In order to do so, such policies rely on the assumption that it is possible to identify such innovations ex ante. The EMInInn project has explored the challenges of identifying whether specific innovations actually generate environmental savings, and it has developed a variety of tools for improving assessments of the environmental consequences of specific innovations. An important conclusion is that it is not always straightforward to identify the environmental consequences that arise from specific innovations, either ex ante or ex post. Unintended consequences are a real risk for technology-specific policies. Detailed assessments of the indirect effects of specific innovations, across the life-cycle and incorporating various economic feedback effects, can help avoid such unintended consequences. It is important to recognise that the environmental consequences arising from the diffusion of a specific innovation depend, in part, on the wider policy conditions. Strong environmental policies, including regulations and emissions pricing policies, can limit the risk that indirect effects result in worse environmental outcomes. In short, eco-innovation policy must be complemented by environmental policy if it is to contribute effectively to environmental improvements at the macrolevel.

1.1 Assessing the “green-ness” of a given good or service is not always straightforward The ultimate test of eco-innovation is whether the development and deployment of new products, processes and business models result in decreases in environmental pressures, relative to a relevant alternative. However, the environmental outcomes associated with the diffusion of a given good or service can be difficult to assess, even ex post, because: • • •

•

Environmental pressures arise across the life-cycle of a product or service. Eco-innovations should be assessed on a cradle-to-grave basis. There are trade-offs between environmental pressures. For example, diesel engines reduce carbon emissions but release more air pollutants per kilometre. Environmental savings from a given eco-innovation may be context-dependent. For example, solar photovoltaics (PV) would certainly be an eco-innovation in Australia (a sunny country with largely fossil-fuel based electricity), but not necessarily in Iceland (a country with low sun and 100% renewable electricity). Shale gas might be considered an ecoinnovation in the short-term if it results in reductions in the use of coal; but probably not in the long-term, since the carbon emissions from widespread natural gas use would breach 2050 carbon targets, unless combined with carbon capture and storage technology. Innovative technologies and processes change over time and as they are diffused. This can change the expected future (or past) environmental pressures associated with those innovations, adding both complexity to the analysis and uncertainty to the outcomes. 4


•

Economic feedbacks can create additional environmental pressures - A major challenge in the assessment of an eco-innovation is the extent to which economic feedbacks are taken into account. Increases in the consumption of a particular product will have effects on other economic actors, supply chains and consumers, resulting in price and income changes. This includes the so-called ‘rebound effect’.

1.2 Eco-innovation policy requires accurate assessments of the environmental consequences of innovations Many innovation and eco-innovation policies target specific technologies or technology fields; as do many environmental policies. Unintended consequences arising from such policies can be reduced with improved technology assessment processes that identify potential indirect effects. Assessment methods used to inform eco-innovation policies do not always assess the full range of potential indirect routes by which unintended environmental effects might emerge. The treatment of economic feedbacks, in particular, is inconsistent in many impact assessments, and typically rather limited in scope. EMInInn has shown clearly that such indirect effects can have significant environmental implications. Of course, very detailed assessments are costly and create administrative burdens, and the degree of detail of assessments should be proportional to the potential impact of the policy change. 1.2.1 Uncertainties should be made clear Assessments about whether an innovation generates net environmental benefits relative to alternatives are subject to uncertainty. Most assessments do not examine all potential feedback and induced effects, and data limits may also prevent detailed assessment. Such uncertainties should be clearly expressed, and information should be provided on how this might affect the outcome of the assessment, for example by conducting sensitivity analysis. 1.2.2

The full environmental consequences of eco-innovation policies depend on wider policy conditions The significance of rebound effects for environmental outcomes depends partly on wider policy conditions. Indirect rebound effects arise when an efficiency innovation saves consumers money, and consumers then purchase other goods and services, which in turn generates environmental pressures. When strong environmental policies—particularly emissions pricing—are in place, those consumer purchases will be directed towards goods and services that have a less damaging environmental profile than would occur in the absence of such policies. EMInInn modelling, using EXIOMOD (Figure 1), has shown how the GHG emissions arising from indirect rebound effects is expected to decline as the carbon price increases.

5


Figure 1. Emissions induced by a marginal increase in European consumption, at different carbon prices

1.3 Indicators for monitoring eco-innovation Most indicators of eco-innovation focus on measuring various activities of innovation systems (such as R&D funding, publications, and patents). These measures are based on judgements about whether particular technologies generate environmental savings when diffused into society. For example, measures based on green patents assume that the patented innovation will generate environmental benefits if it is commercialised and widely adopted. However, judgements about whether any given technology will actually generate environmental savings may not be reliable without detailed assessment. EMInInn analysis has indicated that feedbacks and wider conditions matter. This has implications for attempts to monitor ecoinnovation: methods that rely on indicators of innovation inputs (such as green R&D) and direct outputs (such as green patents) may measure the capacity or potential for eco-innovation—but may not always provide evidence that the innovation process is indeed resulting in environmental improvements. When assessing an eco-innovation policy or action, policy makers should therefore be cautious about using only one indicator as a proxy for eco-innovation, and recognise that it may be misleading. Eco-innovations may also arise from areas of science and technology not thought to be ‘green’, and which as a result lie outside attempts to measure eco-innovation. For example, the technology underpinning the combined cycle gas turbine, which has played a significant role in reducing the environmental impacts of the European power system, has its origins partly in jet engine technology from aircraft. Since eco-innovation is a relative concept, caution is also needed in using indicators based on specific technologies to make comparisons across space and time. The environmental performance of a given technology relative to alternatives – i.e. whether a given innovation is an eco-innovation – depends on the context, since the relevant alternative will differ in different contexts. For example, an indicator based on the diffusion of specific technologies might be unhelpful in comparing 6


countries with very different contexts, since diffusion of those same technologies may result in quite different environmental outcomes—in short, a given technology may be an eco-innovation in one country but not in another.

1.4 Key conclusions Innovation policy is clearly an important strategy for achieving environmental policy goals costeffectively. However, eco-innovation does not necessarily provide an ‘easy way out’. Green innovations that save money are looked upon with great favour by users and policy makers alike because they constitute a win-win option. But it is not always straightforward to identify those innovations, and in any case the environmental wins may be low because of rebound effects. Currently used analytic tools that assess the environmental impacts of specific technologies rarely account for the full range of potential indirect effects. Using improved technology assessment tools, including those developed within EMInInn, may change the assessment of the expected environmental savings associated with specific technological development plans, or policies that are expected to result in the diffusion of particular technologies. In particular, analytic tools that incorporate economic feedback effects and potential life-cycle effects can result in the identification of issues that are sometimes neglected within currently dominant approaches to technology assessment in technology policy formulation. A key point is that the environmental performance of a given technology is not solely an inherent feature of the technology, but depends on wider conditions. In particular, the environmental consequences of ‘rebound effects’ will depend on carbon and energy prices, and policy options to address rebound effects need to be carefully considered. Current policy practice in impact assessment rarely makes such dependencies explicit. It makes sense for technology assessments to make clear where there are such dependencies, and to report a range of potential benefits of an eco-innovation policy package depending on the wider policy context, for example under differing carbon price scenarios. Most fundamentally, however, the analysis makes clear that while ecoinnovation policy is an important part of environmental policy, it needs to complemented by other environmental policy instruments to ensure that it yields its potential environmental benefits. Indicators of innovation system activity, such as patenting, may only provide a partial picture of the trends and drivers of the ultimate contribution of eco-innovation to reductions in environmental pressures. This is partly because such measures rely on an ability to distinguish ‘green’ innovation activities, and partly because of the complexity of the relationship between innovation system activities and environmental outcomes.

7


2 Introduction A note on referencing EMInInn project deliverables This final report from the EMInInn project draws on the previous project deliverables and reports. In order to simplify the referencing, project deliverables are referred to by the deliverable number, i.e. “deliverable 2.2”, thus project deliverable 2.2 is referred to as D2.2. All publicly available project deliverables can be found on the EMInInn website, www.emininn.eu. A full list of project reports and deliverables is found as an annex to this report, in Section 11.

2.1 Fundamentals of EMInInn Eco-innovation is a topic positioned with the highest political relevance. The Europe 2020 strategy and a number of high-level political forums (UNCTAD, UNFCCC, OECD, UNEP, ASEM) have created manifold expectations around the potential of eco-innovation to become the engine to green growth and a prime solution to global environmental pressures such as resource degradation, biodiversity loss and climate destabilization (Montalvo et al., 2011). European policy makers expect that innovation will contribute to reducing the costs of meeting environmental targets, and many see a role for the innovation induced by energy, environmental and climate policy to generate new industries and associated economic benefits (Gallagher et al., 2012, Ekins et al., 2014). Scholars have highlighted the great need to timely and accurately assess presumed “low carbon”, “green” technologies, to avoid supporting technologies that may appear to be ‘clean’ but may in fact turn out not to be (van den Bergh, 2013). Yet currently, policymakers monitor a range of innovation processes often using measures that assume rather than assess the ‘environmental’ credentials of particular innovations (such as ‘environmental R&D’ or ‘clean energy patents’). Hence, available measurement and monitoring policy approaches can fall short to timely and accurately identify such environmental, social, and economic side, unintended, effects. In view of the above, the aspiration of EMInInn is to examine the extent to which the environmental effects associated with specific innovations can be measured at the ‘macro’ level – the total environmental pressures measured at the level of a sector, country, continent or the world as a whole. In this context, different bottom-up and top-down methodologies and indicators were explored and implemented, which will be compared to the extent possible to see whether similar outcomes can be extracted. Further, EMInInn also aims at providing insights on whether an innovation thought to be an eco-innovation actually results in net environmental gains at the macro level, i.e. when system-level and indirect effects are taken into account (including rebound effects and innovation spill-overs). To do so, a comparison between the actual ex post scenario and an exante scenario, or a counterfactual, in which the innovation under assessment would not have existed is proposed (D4.1, p.15). Five economic sectors and a number of corresponding innovations were empirically tested the fields of: energy, ICT, transport, built environment and waste management. The main objective of this report is to present the most important policy messages and recommendations resulting from the EmInInn project. A brief description of the content and main 8


audience for this report is provided below. Since EMInInn has been principally about the measurement and monitoring of the environmental pressures associated with innovation, many of the key policy points concern the way in which policymakers can improve such monitoring and measurement, and why it might be useful to do so.

2.2 EMInInn in relation to the broader eco-innovation literature The main purpose of this section is to place EmInInn in the broader literature and provide an explanation of how it helped to clarify the concept of eco-innovation and to provide methods and tools for a macro-assessment of its environmental impacts. For this purpose, it is important to realise that eco-innovation is a field in the making. Kemp (2010, p. 397) noted that: eco-innovation is a recent concept of which the analytical base is under construction. For the purposes of EMinInn, three critical on-going aspects of this novel field of research can be identified: (i) the absence of an unified general theory of eco-innovation and the availability of manifold bodies of literature informing it; (ii)the degree of fragmentation of its epistemic community, and (iii) the need for integration/complementarities of methods and indicators to assess the real net environmental benefits of innovations. The rich variety of bodies of literature and associated theories informing eco-innovation poses challenges to be regarded as a generally accepted theory of eco-innovation research (c.f. van den Bergh et al 2011). Rennings (1998), Ekins (2010) and Kemp (2010) provided an overview of bodies of literature, and concepts around eco-innovation. These author informed their reviews with insights from the fields of evolutionary economics, sustainable development, innovation studies and ecological economics. At first glance, the latter fields can be assumed as being an integral component of the theoretical underpinnings of eco-innovation studies. In the opening article of the Environmental Innovation and Societal Transitions Journal, Van den Bergh et al (2011, p.3) identified five disciplines (which can be subsequently subdivided into eleven sub-disciplines) involved in ecoinnovation studies: (1) management studies (business, management and accounting); (2) environmental studies and transport studies (environmental sciences), (3) geography and political sciences (social sciences), (4) ecological economics, evolutionary economics (innovation studies, environmental economics), and science and technology studies (economics, econometrics and finance), and (5) energy studies (energy). Diaz Lopez (forthcoming) identify the following additional areas around eco-innovation research: behavioural sciences, sociology (environmental sociology, industrial sociology), engineering (design, chemical, materials), computer sciences, chemistry, and agricultural and biological sciences. Until recently, there has not been broadly integrated a unified, coherent community devoted to the study of eco-innovation. 1 Notwithstanding, the rich set of bodies of literature provide complementary elements to bring about suitable and pertinent research in this field of knowledge. The challenge lies in improving communication and collaboration across different epistemic community. The study of eco-innovation has been hitherto performed by a rather fragmented community of scholars (c.f. van den Bergh et al 2011). The growing academic interest in the study of eco1

The recently-funded “European Global Transition Network on Eco-Innovation, Green Economy and Sustainable Development”, funded by the EC’s H2020 as a coordination and support action, is hoped to provide a boost to the creation of an epistemic community of eco-innovation scholars. Most partners in EMinInn are part of it and currently share a number of scientific activities devoted to different angles of the study of eco-innovation.

9


innovation can be illustrated with the results of bibliographic data (Arundel et al., 2007). Google scholar reports 17.490 documents including the exact phrases ‘eco-innovation’ and ‘environmental innovation’ in the field “everywhere in the article” and 908 documents in the field “in the title of the article”. The period 2010-2014 concentrates 74% of publications with the word ‘eco-innovation’ in the title of the document.2 A search in the Scopus databases encountered that 120 journals have published at least one paper with the words ‘eco-innovation” or “environmental innovation” in the TITLE field in the period 1991-2014. With 208 published manuscripts, such figure represents ratio of 1,7 articles per available journal. 3 Such fragmentation imposes methodological and empirical challenges to any multidisciplinary attempt to study of eco-innovation, such as in the case of EMinInn. In view of such of descentralisation of knowledge, it is not illogical to think that scholars in different communities may not sufficiently inform one another about useful theories, methods and tools that could complement their views about eco-innovation and its contribution to sustainability. Determining what innovations bring real environmental and social benefits, and identifying critical conditions for the occurrence of benefits, is one of the most challenging avenues of research (c.f. van den Bergh 2013). This situation is aggravated by the insufficient agreement about intentionality, scope and degree of change, novelty, contribution to environmental sustainability and resource efficiency, effects of barriers and drivers and effective policy mixes across the life cycle of ecoinnovations (Ekins, 2010, van den Bergh et al., 2011). Different disciplines have attempted to provide an account of the resource use, environmental degradation and pollution effects without sufficiently informed one another. Some examples are provided below: •

•

In the field of economic history of innovation, Rosenberg (1976) was one of the first scholars discussing the role of innovation and technological change in addressing environmental challenges. The determinants of eco-innovation have been studied by innovation scholars such as Rennings, Horbach and Johnstone, but the macro-environmental outcomes were not a topic for investigation. Likewise, in the field of transition studies, the greater sustainability of new systems of mobility, energy and food production the macro-environmental benefits are not a topic of research. In the field of environmental and sustainability sciences, environmental aspects are studied with respect to particular ‘functional units’.

2

Search performed on the 23.12.2014 using the advance search function of google scholar. The time span was set to; (1) all years and (2) the years 2010 to 2014. The search excluded the categories: citations and patents. 3 Search performed on the 23.12.2014 using the advanced search function of Scopus database. The search categories were set to ‘journals’ only, and excluded the results outside the categories of ‘articles, articles in press, reviews’. The resulting figure was 208 documents, comprising 183 published articles, 17 reviews and 8 articles in press. The top 5 journals for eco-innovation scholarly work reported by this search method were: the Journal of Cleaner Production (20), Ecological Economics (9), Business Strategy and the Environment (8), Research Policy (8) and Industry and Innovation (7). The launching of the Environmental Innovation and Societal Transitions Journal offers an outlet for eco-innovation publications but it has been rather unsuccessful in attracting those publications. One reason for this is that it does not yet have an official impact factor (which depends on meeting very strict requirements about review and citations). From 102 manuscripts in this Journal (articles, articles in press, reviews) only 14 documents include the word ‘eco-innovation’ anywhere in the text, and 6 documents include the word ‘environmental innovation’ in the fields ‘title, abstract and keywords’. Not a single article has included the word ‘eco-innovation’ as part of the title, albeit 2 articles have used it in the keywords. This publication published 1 special around diffusion of environmental innovation in the year 2014, with 2 original articles and 1 introduction to the special issue explicitly addressing the topic of diffusion of environmental innovation in the title of the document.

10


•

•

The area of design engineering was among the first in dedicating full attention to the minimization of environmental impact from the design stage of product innovation in the late 1990s, incorporating such concerns into early eco-design methodologies (c.f. Johansson and Magnusson, 1998, Jones and Harrison, 2000). Using eco-design methods it possible to estimate an eco-efficiency potential of products or technologies (c.f. Kobayashi et al., 2005). The support tools for performing such a task are directly linked to the life cycle tools referred to above. More recently, the field of environmental and resource economics have also developed methods to assess macro-environmental impacts of technologies, e.g. MFA and environmentally extended input-output methods (c.f. Tukker et al., 2009).

EMinInn was set to provide a common understanding on eco-innovation across different scientific disciplines, including sustainability and environmental sciences, innovation studies, evolutionary economics, behavioural economics, environmental and resource economics, industrial ecology and different sub-disciplines of engineering. EMinInn also aims at creating common definitions and analytical boundaries, and to develop integrative methods to assess the macro-environmental impacts of innovation. Judgements about the environmental effects of applying a technology, and whether these are more or less intense than those of applying a relevant alternative, are typically made at a ‘micro’ level. The system boundary is typically drawn rather narrowly around the technological system or artefact, and wider systemic effects are ignored. This means that it is difficult to know whether scaling up such assessments to the macro-level makes sense, or whether other factors will undermine the relevance of micro-level assessments (D4.1 p. 15). EMinInn noted that when dealing with specific innovations (as opposed to innovation as a general process), the environmental impact of an innovation depends on aspects of use and induced effects. In applying metrics, one should also look at behaviour and dynamic effects. A relevant issue in this respect is what is done with the old product. For example, when people buy a more energy-efficient refrigerator, what do they do with the old one: Is it scrapped, or do they keep it? If the innovation helps to save money, one should look at the impact of induced expenditures.

2.3 This report This reports aims to summarise the way that the tools developed within EMInInn can support policymakers by improving technology assessment procedures in order to make policy making to be more sensitive to and aware of wider environmental effects. Key questions addressed in this report are: a) To provide an overview of the location of Emininn in the broader eco-innovation literature; to offer a brief summary of concepts and theories b) To present an overview of methods, indicators to evaluate the environmental effects of supposed eco-innovations c) To attempt to explain to what extent can the final outputs of this project bring greater clarity for policymakers seeking guidance in the area of methods and indicators about the environmental impact of innovations. This report is primarily written for EU policy makers preparing policies and policy evaluations in the areas of innovation, climate, energy, resource efficiency and sustainability more broadly.

11


2.4 Outline The content of this report is divided in the following sections. Following this introductory section, the remainder of the document discusses some key issues on what eco-innovation is, how it can be measured, and the relationship between innovation, policy, and environmental performance (section 3). Section 4 then provides an overview of the way in which policymakers currently take account of the environmental effects of eco-innovation in the policy cycle, through a focus on the methods and indicators used in impact assessments, monitoring and ex-post evaluations. Sections 5 and 6 outline what we have done in the analytic work packages for the development and application of methods for assessing, both ex post and ex ante, environmental impact of specific innovations and the methods for monitoring eco-innovation at the macro-level. Section 7 summarises the results arising from EMInInn on the conditions under which innovation processes generate environmental benefits. Section 8 summarises the specific results of case studies and the macro-modelling in terms of policy making support, while section 9 draws conclusions.

12


3

Concepts and analytical model to understand the environmental impacts of eco-innovation

3.1 Concepts This section provides key definitions in the realm of innovation, eco-innovation and its environmental effects. It subsequently outlines an analytical model that illustrates how innovation affects the environment. 3.1.1

Definition of eco-innovation

Innovation should be distinguished from invention and technological change (c.f. Rosenberg, 1982). Invention refers to discovery, whereas innovation refers to technologically novel or improved material goods, intangible services or ways of producing goods and services (Edquist, 2005). Kemp and Foxon (Kemp and Foxon, 2007) defined innovation as the production, assimilation or exploitation of a novelty in products, production processes, services or in management and business methods. The latter is a definition following the concepts and categories defined by the Oslo Manual (OECD and EUROSTAT, 2005) (see below). Eco-innovations can be broadly defined as those innovations that contribute to the environmental dimension of sustainable development (Rennings, 2000a). Within the context of the ECODRIVE project, Huppes et al. (2008) suggested that eco-innovation should be understood as the combined improvement of economic and environmental performance of society. Within the context of the MEI project Kemp and Pearson (2008, p. 7) defined eco-innovation as the “the production application or exploitation of a good, service, production process, organisational structure, or management or business method that is novel to the firm or user and which results, throughout its life cycle, in a reduction of environmental risk, pollution and the negative impacts of resources use (including energy use) compared to relevant alternatives.” The European Commission (2011p. 2) defines ecoinnovation as: any form of innovation resulting in or aiming at significant and demonstrable progress towards the goal of sustainable development, through reducing impacts on the environment, enhancing resilience to environmental pressures, or achieving a more efficient and responsible use of natural resources. The term eco-innovation is often used interchangeably with that of environmental innovation, especially within the field of environmental and resource economics. 4 Based on the results from the Ecodrive project, Ekins (2010) provides a useful distinction between environmental innovations and eco-innovations, being the latter term associated with significant gains in both economic and environmental performance. According to this author, environmental innovations enable relative environment and/or the economy improvements. In other words, these are innovations where improved environmental performance is accompanied by a share of deteriorating economic performance and vice versa (Ekins, 2010, p. 270). 4

Other terms associated to eco-innovation are: environmental technology, low carbon innovation, green innovation, environmentally friendly innovation, sustainable innovation, ecological innovation, clean technology, green technology, cleantech innovation, smart innovation, etc. While their use for practical purposes it is commonly accepted as a valid practice in scholarly work, for analytical purposes it is required to provide an adequate differentiation of the real meaning, semantics, epistemological stand and field of application of each concept.

13


A basic challenge of EMInInn is the difficulty of identifying whether a given innovation is, or is not, ‘truly green’. As noted in the introduction, there is a great need to address the theoretical and empirical challenges for measuring and monitoring eco-innovation and eco-innovation processes. Van den Bergh (2013) argues that reliance on a clear distinction between ‘clean’ and ‘dirty’ technologies is problematic because of many of the indirect effects that have been examined in the EMInInn project. For example in the case of energy innovation, this author highlights the significant energy and material inputs required for many renewable energy sources, citing bioenergy and wind as examples. For this reason, within EMinInn an effort has been made to characterise ‘supposed’ eco-innovations processes from ‘true’ eco-innovations resulting in positive environmental improvements at the macro-level. A salient aspect of the methodological approach used by EMinInn was the use of appropriate categories for the selection of relevant innovations and eco-innovations. The following section present an overview of the literature in this regard. It then presents a summary table with the main technologies addressed in the empirical validation of EMinInn work. 3.1.2

(Eco)innovation typologies

Various authors have developed typologies of different general kinds of innovations. These categorise innovations according to their domain. One of the best known examples is provided in the Oslo Manual (OECD and EUROSTAT, 2005), prescribing the well-known categories of product, process, organizational and marketing innovation. Innovation classifications also note their relationship with incumbent ways of doing things: incremental vs. radical, sustaining vs. disruptive; etc. (Abernathy and Clark, 1985, Bower and Christensen, 1995); and finally there are innovations that are recognised to have effects across a range of sectors, which are often so-called ‘general purpose technologies’ that have a wide scope of application (Bresnahan and Trajtenberg, 1995). Recent classifications of innovation take into account the potential to disrupt or change entire socio-technical systems (technology and user practices, markets and institutions): incremental, social, transformative or techno-fix innovation (Kemp, 2011). These are ‘system innovations’ inducing large scale transformations to the way socio-technical functions are fulfilled (Weterings et al., 1997) (see following section). There are also typologies of types of environmental innovations. Early classifications of ecoinnovation were technology-oriented (e.g. in Bartolomeo et al., 2003, Kemp, 1997); based on the early taxonomies of environmental and cleaner technologies elaborated by the OECD (1985), ACOST (1992) and Skea (1995). Sub-categories of the latter typologies include: waste management, recycling, waste minimisation, clean technology, measurement & monitoring equipment and clean products. Newer classifications combine general kinds of innovations with different sub-categories of environmental technologies. For example, as a result of the MEI project, Kemp and Pearson (2007) use four groups of eco-innovations: environmental technologies, organisational innovations for the environment, product and service innovations, and green systems innovations 5. 5

Alternative systems of production and consumption that are more environmentally benign than existing systems.

14


For the purposes of EMInInn definitions and classifications focusing on products and technologies, process and organisational and marketing decisions of companies were incorporated albeit these were considered somehow limited. As it will noted in subsequent sections, the scope of this project was also interested in systemic eco-innovations involving behavioural and society-level change (e.g. transport systems of inter-modality). The focus of EmInInn was therefore on eco-innovations that produce results on the ground; but in this project motivations and institutional mechanism behind such results, especially the role of public policies, were also considered. An overview of the classifications created throughout EMinInn case studies in reference to and selected innovations is provided in Table 1. In the EMInInn project we were interested in the diffusion of innovations because the impact will depend on the extent to which it is used by people and organisations. The diffusion depends on the extent to which the innovation confers benefits on its user, which in turn depends not only on the techno-economic characteristics of the innovation but also on framework conditions. Diffusion is governed by in the innovation and changes in the adopter environment such as the new environmental regulation, aging of the capital stock, the obsolescence of products, and changes in the preferences and life circumstances of people (D2.2, p. 44). The following section expands on the topic of systemic aspects of innovation and eco-innovation. Table 1 Categories of eco-innovation and selected cases in EMinInn

Work package WP4- Energy Energy innovation based on attributes of the energy system

Eco-innovation classification Resource innovations

WP5-ICT

Conversion and supply-chain technologies End of pipe, waste disposal and clean up innovations System innovations Energy-consuming product and process innovations Energy demand efficiency innovations Front end

Sub-systems of the internet system

Access networks Back end Applications

Selected cases ex-post assessment • Wind energy • Solar PV •

• •

Combined cycle gas turbines (CCGT) Flue-gas desulphurisation (FGD) NA NA

• • • • • •

•

• WP6- Transport Transport technologies with a high micro-level relative environmental

… framed by favorable conditions.

…framed by unfavorable conditions.

• • • • • •

Selected cases ex-ante assessment • Onshore wind • Offshore wind • PV panels • NA • • •

Carbon capture and storage (CCS) NA NA

NA

•

NA

Mobile phone Smart phone Wireless networks Data centres Electronic product reuse Long distance tourism

•

NA

• • •

Catalytic converters Diesel engines* Direct fuel injection (DFI) systems* High speed rail systems Battery electric vehicles (BEV) Fuel cell vehicles (FCV)

•

NA NA Electronic retail trade Electronic wholesale trade NA

•

• •

Full battery electric passenger cars Hydrogen fuel cell

15


performance …

Transport innovations…

WP7 - Built environment Energy efficient housing based on the concept of ‘trias energetica’

… and preliminary evidence (e.g. changes in consumption factors) of notable impact of rebound effects. …that have been responsible or have the potential for substantial organizational changes to transport systems. Reduction of the demand for energy

• • •

Diesel engines* DFI systems* High speed rail systems

•

passenger cars NA

•

Park-and-ride (P+R) facilities Car sharing Public bike schemes Insulation of floor, wall and roof / loft Multiple-glazed windows District heating Ground source heat pumps High efficiency (eco) boilers Energy saving light bulbs / LED’s / Fluorescent lights NA Material recycling and composting Incineration and energy recovery Landfill

•

NA

•

Natural Fibre insulation Triple Glazing

• • • •

Sustainable and renewable sources of energy

• •

Most efficient use nonrenewable energy

• •

WP8 – Waste Waste management according to the EU ‘waste hierarchy’

Prevention Reuse, recycling

• •

Incineration and energy recovery

•

Landfill

•

• • • •

Solar Thermal Photovoltaic (PV) panels Mechanical Ventilation with heat recovering

• • •

NA Recycling Composting

•

Incineration

•

Landfill

Notes: NA= category not considered or assessed Source: elaborated based on data and information in Deliverables D4.1, D5.2, D6.1, D71, D8.3 and D9.2 of EMinInn

3.1.3

A systemic view of (eco)innovation

The diffusion of eco-innovations is not just a matter of information transfer or of environmental regulations becoming stricter. Whether an adopter will opt to buy and use a particular environmental technology at a given moment in time will depend on the (perceived) costs and benefits of adopting it, which in turn depend on wider system aspects. Systemic innovations typically diffuse slowly, because the diffusion is associated with changes in infrastructures, institutions and economic prices and conditions. For macro-environmental assessment, it is possible to distinguish between: (i) systemic change induced by innovation processes and (ii) innovation and the framework conditions in the innovation system. In relation to system level change, Weterings et al. (1997) identified three levels of environmental innovations: system optimisation, partial system redesign and function innovation of technological systems: • Optimisation focuses on improving existing products, processes or infrastructure. The concern is to modify systems which already have a commercial use. • Redesign, the actual design of existing products, processes or infrastructure is partly changed. Specific features of the system are changed, for instance by choosing to use materials that can be made suitable for reuse in the disposal stage.

16


•

System innovation develops a new system, which perform the same function better. This can result in a radical change as to how the function is fulfilled.

Weterings et al (2007) contend that system optimisation offers the greatest environmental gains in the short term and system innovation the biggest gains in the long-term. Hypothesised time patterns for environmental improvement are shown in Figure 1. While this view has some plausibility, it is not backed by empirical studies.

Figure 2. Environmental innovation in view of three levels of system change and environmental improvement by factors Source: Weterings et al. (1997)

In relation to the innovation systems focus, useful theories to understand the broader context and generative mechanisms of innovation are: innovation systems (systems of innovation, sectoral innovation systems) and sustainability transitions (TIS, multi-level perspective) (Coenen and Díaz López, 2010). A brief description is provided below. •

•

•

Systems of innovation comprise actors and technologies, the networks through which they interact, and the institutions that govern those interactions (c.f. Edquist, 1997). Freeman (1995) defined the innovation system (IS) as: The network of institutions in the public and private sectors whose activities and interactions initiate, import, modify, and diffuse new technologies. Sectoral innovation systems (SIS) is a useful framework to characterise patterns of innovative activity of specific industries (c.f. Pavitt, 1984). In this approach, innovation takes place at the level of products, technologies, firms, and institutions and is caused by the heterogeneity among actors, often constrained by market selection mechanisms and lock-in effects (Malerba, 2004). The technological innovation systems (TIS) approach (Hekkert et al., 2007, Bergek et al., 2008) sees innovation as the result of processes of alignment between different components of a socio-technical system around a technology.

17


•

•

The multi-level perspective (MLP) based on the distinction between niches, regimes and landscape is useful (Geels, 2002). In this approach an emerging technology will find initial use in niches, special application domains, inside or outside a regime, whose evolution is helped by certain landscape developments but obstructed by others (D2.2, p. 48). Lastly, key theoretical underpinnings of regional innovation systems have been recently used to explain the territorial nature of sustainability transitions, the so-called transition regions (TR). La latter approach unavoidable show a large degree of ‘natural’ variety in institutional conditions, networks, actor strategies and resources across territorial scales (Coenen et al., 2012). 6

A key accepted characteristic of eco-innovation is the so-called double (market) externality problem, involving an environmental and a knowledge-sharing externality (Rennings, 2000a). Market failures are evident when facing the challenge of addressing a negative economic externality such as air pollution. In addition, technological change also faces a public good nature condition that hampers its wider diffusion (Rennings, 2000a, Del Río et al., 2010). In addition to market imperfections, ecoinnovation scholars have acknowledged the presence of failures at the level of the system (Kemp, 2011). The systems failure framework (SFF) makes it possible to make a structural analysis of a sector, and to assess which factors hinder or drive innovation in that system. SFF gives a balanced account of the conditions surrounding innovation (e.g. government policies, cultural differences), entrepreneurial actions (e.g. industry actions, introduction of new products) and market functioning (e.g. demand, prices) (Klein Woolthuis et al., 2005, Klein Woolthuis, 2010). Within the context of EMinInn, the system failure framework was used in the analytical work package of the built environment. Within the context of EMinInn, all analytical work packages used (or made reference to) the theories hitherto presented. An overview of the underpinning theories used to inform EMinInn’s empirical work is included in the table below. Table 2 Theoretical frameworks guiding the case studies of EMinInn

EMinInn work package – case studies WP4- Innovations in the energy system WP5 -The internet system WP6 - Transport innovations WP7 - Energy efficient buildings WP8 - Waste management

IS Used Used Used Used Used

SIS Used Used Used

Systems approaches to innovation TIS MLP TR Used Referenced Used Referenced

SFF

Used

Source: elaborated based on data and information in Deliverables D4.1, D5.2, D6.1, D7.1, 8.1 and D8.3 of EMinInn

It is acknowledged in the academic literature that barriers to eco-innovation may be understood as stemming from market failures and system failures acting in them. The views described above are closely related to the identification of drivers and barriers to (eco)innovation; a topic which is briefly discussed in the following section.

6

Deliverable D2.2 of EMinNN provide an overview of these approaches; alternatively refer to Coenen and Diaz Lopez (2010).

18


3.1.4

Drivers and barriers of (eco) innovation

Barriers to (eco)innovation can be defined as a factor exerting a negative impact on the (eco)innovation process (Piatier, 1984). Factors with a positive influence can be called drivers, stimuli or facilitators (Hadjimanolis, 2003). The perceived effect of them tend to vary depending on the type of innovation, the size of the firm, the sector, the geographical location, the business cycle and the stage of development of the (national or technological) innovation system (Hadjimanolis, 2003). For analytical purposes barriers are often used as a separate category to drivers to innovation (Hadjimanolis, 2003). But it is important to note that the intrinsic relation between the two of them. On the one hand barriers often emerge because of the absence of facilitating, enabling factors driving innovation and change. On the other hand drivers may exert the opposite effect along different stages of the innovation cycle, depending on changing conditions in markets, sectors or technology (Koberg et al., 1996, Hadjimanolis, 2003). Drivers and barriers can be internal (endogenous) or external (exogenous) to the firm (Piatier, 1984). Internal barriers can be found in relation to resources or capabilities for innovation, absorptive capacity, willingness or resistance to innovation and change, etc. External drivers and barriers are often referred to those pertaining to the environment surrounding the innovation (Michaelis and Carey, 1973). Examples of internal drivers/barriers to eco-innovation are the presence of absence of environmental responsibility (green ethos) and environmental management systems. External drivers/barriers are positive or negative effects from environmental regulations, innovation subsidies, pollution taxes, demand from users, and pressure from local communities (D2.2, p. 24). Drivers and barriers to eco-innovation can also be classified in the following categories: institutional, market, economic and finance, behavioural, technological and organisational (Michaelis and Carey, 1973). Similar categories can be identified in the eco-innovation literature (c.f. Montalvo, 2008, Bastein et al., 2014, Weber, 1997). A considerable amount of (empirical) work aiming at a better understanding of drivers and barriers to eco-innovation is available for a number of OECD and European Union countries (c.f. Klemmer et al., 1999, Triguero et al., 2013, Horbach et al., 2012, Mazzanti and Zoboli, 2009, Frondel et al., 2007, Kesidou and Demirel, 2012, Ashford, 1993). However, providing a complete overview of the effect of internal and drivers would be a major task beyond the objectives of this report. From the literature the following insights have emerged: that eco-innovation requires effort and money, that the creation of an eco-innovation requires capabilities to eco-innovate and willingness and cooperation with other knowledge actors when part of the knowledge needed to innovate is not available internally. The willingness may stem from various sources: market-based pressures for cost-reduction, commercialisation prospects (demand from green customers), pressures for regulation, NGOs, clients, the parent company. For their eco-innovation, companies may rely heavily on external knowledge actors from which they source knowledge or technology, or on internal actors. Decisions to eco-innovate generally depends on internal judgment about capabilities, a green ethos, and managerial expectations about gains to be had and the costs and risks involved. Companies that obtained positive results with eco-innovation in the past can be expected to be more inclined to eco-innovate. Like normal innovation, eco-innovation depends on internal champions and wider system aspects such as the availability of finance, the presence of knowledge holders possessing relevant knowledge and a (regional) culture for cooperation. The ease with 19


which an innovation can be protected to imitation may affect innovation decisions but there is little evidence of this working as a negative factor to innovation; it is primarily relevant for the choice of strategy for protection (secrecy, patents, using marketing to obtain a strong position in the market). In the context of EMinInn the relationship between eco-innovation and green growth was examined paying attention to internal and external stimuli and hampering factors affecting innovation diffusion. For this reason deliverable 2.2 offers an overview of drivers and barriers based on a review of the literature. Particular attention was paid to show known differences depending on the type of innovation, the economic sector of reference, actor in the innovation processes (adopters, developers, and regulators), and across EU Member States. For example, it was noted that Horbach et al (2012)) found that current and expected government regulations are particularly important for pushing firms to reduce air (e.g. CO2, SO2 or NOx) as well as water or noise emissions, avoid hazardous substances and increase recyclability of products. Cost savings are found to be an important motivation for reducing energy and material use, pointing to the role of energy and raw materials prices as well as taxation as drivers for eco-innovation. Customer requirements are particularly important for green product innovation and for process innovations that increase material efficiency, reduce energy consumption and waste and the use of dangerous substances (Horbach et al., 2011). D2.2 also provides a brief snapshot analysis of eco-innovation barriers between EU27 and EU15 countries using data from the Eurobarometer survey and the Ecoinnovation Observatory. 7 Overall, it is reported that the new member states encounter more challenges to perform well in eco-innovations. Next to financing (lack of funds within the enterprise), market uncertainty and lack of access to information and technology support are encountered as major barriers. All case studies in EMinInn also performed a literature review of known effects of drivers and barriers to eco-innovation. 8 Detailed empirical work on the drivers of eco-innovation has been undertaken for the waste sector in WP8, using econometric analysis. Results are reported in Chapter 5 of this report. While the identification of drivers for innovation is highly regarded as a basic ingredient for upscaling and diffusion, the identification of barriers are key to design of effective policy interventions. The role of policy and current policy responses in relation to eco-innovation are briefly discussed below. 3.1.5

Policy responses

As noted in the introduction, policy makers expect that more innovation will contribute to reducing the costs of meeting climate and sustainability targets, and many see a role for the innovation induced by energy, environmental and climate policy to generate new industries and associated economic benefits. Opposite views on the role of innovation (diffusion) as a solution to climate and ecological problems and government interventions have been signposted by scholars. On the one hand, it has been acknowledged that a transition to global sustainability requires the international 7

For example, the Eurobarometer survey includes barriers to eco-innovation such as: lack of funds within the enterprise, insufficient access to existing subsidies and fiscal incentives, technical and technological lock-ins (e.g. old technical infrastructures), lack of external financing, existing regulations and structures not providing incentives to eco-innovate, etc. 8 The review of drivers and barriers to eco-innovation was an important component of the analysis of the environmental effects of the diffusion of past innovations. Refer to deliverables D4.1 (energy), D5.2 (ICT), D 7.1 (built environment) and D8.3 (waste).

20


diffusion of eco-innovations and the creation and expansion of lead markets (Rennings, 2014). On the other hand, technological innovation is only regarded as a modest part of the solution to sustainability. Given the slow dynamics of development and diffusion of innovation and the high pace of ecological and climate degradation, better forms of environmental regulations are required if we are to significantly tackle GHG emissions and other critical factors (van den Bergh, 2013). It is clear that policy makers and scholars agree on the need of appropriate policy intervention. From an innovation policy standpoint, policy intervention is justified when two conditions are met: a problem must exist and relevant government agencies must be able and have the capacity to address it (Edquist, 2005). In this, the identification and removal of market and system failures is a core area of policy action (c.f. Coenen and Diaz Lopez, 2010). The policy rationale for the support of eco-innovation rests on the assumption that market mechanisms will fail to deliver the expected support for their uptake (Kemp, 2011). As a result, eco-innovation can be seen as a policy objective on its own – e.g. for the creation of new green ventures and the general uptake of the eco-industry. And/or it can be seen as a mechanism to achieve objectives of sustainable development, for example when requiring solutions to an environmental problem such as soil degradation or sulphur emissions (Diaz Lopez, et. al. forthcoming). Typical examples of environmental and innovation policy responses to promote eco-innovation are: subsidies for R&D and innovation projects, financial support for formation of university-industry collaboration networks, establishment of regulations and standards, support to business intermediaries such as business incubators, provision of technological information and support, etc. (c.f. Ekins 2010; Kemp, 2011). Policy responses are heavily influenced by the context which it seeks to change, the problem being identified, and the definition of policy targets. Innovation policy has traditionally paid attention to the public good of innovation vs. private costs to consumers, green taxation and environmental regulation (Kemp and Pontoglio, 2008). Environmental policy has observed that innovation suffers from two market failures – the public good nature of knowledge and non-internalisation of externalities (Rennings, 2000b, Popp, 2010), calling for policies of push and pull. In view of the above, there is a consensus among eco-innovation experts that eco-innovation development and diffusion requires a mix of policies. Ekins observes that it is increasingly common to seek to deploy policy instruments in optimal policy ‘mixes’ or ‘packages’, in order to enhance their effectiveness across the pillars of sustainable development (Ekins, 2010). Whilst the need for policy mixes is well-understood, the precise nature of them has to be determined on a case by case basis, raising difficult questions about the coordination and timing across the innovation cycle (Coenen and Díaz López, 2010). According to van den Bergh (2013) complementary, timely and relevant policies are the only way to avoid paradoxes, leakages and counterproductive outcomes when combining environmental-climate—energy-innovation policy (van den Bergh, 2013). This author proposes that much greater attention needs to be paid to assessment of environmental and energy technologies, to avoid supporting technologies that may appear to be ‘clean’ but may in fact turn out not to be; second he argues that, regardless of efforts to identify and support cleaner technologies, emissions pricing is critical (ideally through a cap-and-trade system) to ensuring that the substitution and spending decisions made by consumers and producers are steered towards activities with lower emissions (van den Bergh, 2013, van den Bergh, 2011). This author argues that only such a combined approach can render long term, positive, environmental results – avoiding rebound effects and the green paradox in the forthcoming period of the next 30 to 50 years. 21


In the context of EMinInn, Kemp et al. (2013) noted that a transition to sustainability facilitated by public support to eco-innovation may not be straightforward. 9 The underlying reason is threefold: environmental challenges are many and they compete with other challenges, such as gender equality, energy sufficiency, etc. Second, markets favour innovations requiring few changes in systems of provision and lifestyles of people. And third, information problems and problems of acceptance heavily constrain the set of possible government interventions. The case of electric mobility provides a useful illustration of the above. This case shows that policy should be concerned with green power and recycling, car-restraining policies for conventional vehicles if e.g. electric mobility is to produce a major reduction in carbon emissions. This suggests that the challenge for policy is to stimulate eco-innovations, to timely identify its (undesired) impacts and to maximize its environmental benefits. The latter challenges may actually be far more difficult than the first one, as it is likely to involve the use of unpopular measures (D2.2, p. 80). In practice, policymakers monitor a range of innovation processes, often using measures that assume rather than assess the ‘environmental’ credentials of a particular innovation or innovation process (such as ‘environmental R&D’ or ‘clean energy patents’). The empirical results of EMinInn showed that identifying whether and how (innovation) policies generate ‘true’ eco-innovations is possible. However, such process constitutes a rather complex task often requiring data and indicators which are not readily available. In the context of available methods and indicators is that the following section presents a review of the state of the art in eco-innovation measurement.

3.2 Measuring eco-innovation: progress and challenges This section describes possible measures for eco-innovation, discusses data availability and offers suggestions for use. Eco-innovation can be measured on the basis of four indicator categories (Arundel et al., 2009) drawing on Arc and Audretsch (1993) and Kleinknecht et al. (2002)): •

• • •

Input measures: Research and development (R&D) expenditures, R&D personnel, and innovation expenditures (including investment in intangibles such as design expenditures and software and marketing costs); Intermediate output measures: the number of patents; numbers and types of scientific publications, etc; Direct output measures: the number of innovations, descriptions of individual innovations, data on sales of new products, etc; Indirect innovation indicators derived from aggregate environmental performance data.

2.2.1 Input measures 9

Deliverable D2.2. contains a section on polity rationale for eco-innovation. This chapter starts by reviewing relevant policy considerations around green growth and sustainable development. It then provides an overview of different analytical tools for assessing the relevance of supporting eco-innovation, including those related to policy mixes.

22


Input measures are measures about the inputs for innovation: Research and development (R&D) expenditures, R&D personnel, and innovation expenditures. R&D statistics are widely used in innovation research as a measure for innovation activity. One reason for their popularity is their availability. R&D data are collected by national bureaus and made available in aggregate form. A limitation of R&D is that they capture formal R&D, typically within formal R&D laboratories, and underestimate R&D conducted by smaller firms, which is often done on a more informal basis (Kleinknecht et al, 2002). Also R&D does cover nontechnological innovation such as marketing, organisational and institutional innovations and cannot capture the efforts of service sectors. The OECD's ANBERD database (www.oecd.org/sti/anberd) presents annual data on industrial R&D expenditures. Series are published from 1987 to 2010 for 32 OECD countries and 6 non-member economies. Another interesting database is the STAN database offering information about R&D expenditures and the value of embodied R&D. 10 From the GBAORD data base, information is available about government budgetary appropriations or outlays for R&D for control and care of the environment. In the Community Innovation Survey inquires into whether companies are doing R&D allowing for a comparison across sectors and countries. R&D data for environmental innovations are not systematically collected on an ongoing basis and it is recommended that will be done in the future. 2.2.2 Intermediate output measures These consist of patents and scientific publication and citations. Patents are the most commonly used indicator for inventions (Dodgson and Hinze, 2000, p. 103). A patent is an exclusive right to exploit (make, use, sell, or import) an invention over a limited period of time (20 years from filing) within the country where the application is made (OECD 2004, p.8). Patents are granted for inventions which are novel, inventive, and have an industrial application, but patents do not need to be commercially applied. Patents have several advantages over R&D expenditures: (i) they explicitly give an indication of inventive output, (ii) they can be disaggregated by technology group, and (iii) they combine detail and coverage of technologies (Lanjouw et al., 1998). Moreover, they are based on an objective and slowly changing standard because they are granted on the basis of novelty and utility (Griliches, 1990). Patent counts can be used as an indicator of the level of innovative activity in the environmental domain. We should note that the value distribution of patents is highly skewed. A few patents are commercially valuable whereas the majority have little value. 10

See http://www.worldklems.net/conferences/worldklems2010_Webb.ppt#282,2,The STAN family

23


Hence the usefulness of simple patent counts is limited, as they give equal weight to patents of very different values. Citation analysis can be used to select relevant patents and eliminate patents that have no commercial application. An important new development is the EPO/OECD PATSTAT database, a new database, containing 60 million patent applications from over 80 national and regional patent offices, going back as far as the 1880s in some cases (Johnstone and Hascic, 2008, p. 8). In this database not only inventions in end-of-pipe technologies but also inventions in “more integrated technological innovations” (such as fuel cells for motor vehicles) may be identified according to Johnstone and Hascic (2008, p. 8). Patent data can be used to estimate and compare inventive activity for different technologies within a country and across nations and per environmental medium Figure 2.

Figure 3. Number of patent applications for “environmental technologies” for environmental medium. Source: Johnstone and Hascic (2009, p 9)

Given their availability for nations and products (technologies), patent data are an interesting source of information. It means we have macro-data and micro-data that can be used. Patents may also be used to measure international technology diffusion. Patent applications at multiple countries are indicative of international diffusion of the knowledge contained in the patent (Eaton and Kortum, 1996, p.254). Filing a patent application in a given country is a signal that the inventor expects the invention to be profitable in that country, which is seen as a potential market. Thus, in principle, researchers can use data on multiple filings of patents (patent families) to track technology diffusion across countries (Oltra et al. 2009, p. 137). From the point of view of assessing impacts, the data are too 24


crude to be of much use. It is better to use data on product sales for measuring diffusion. Patents can be used in prospective analysis but innovation diffusion prediction should not be based on patents alone. 2.2.3. Direct output measures Direct output measures look at the occurrence of innovations and the characteristics of the innovation. Two approaches may be used: an object–based approach and a subject-based approach. In a subject-based approach, companies are asked whether they have a adopted or developed an innovation of (a specific) environmental benefit. In an object-based approach, details are collected on a real existing innovation (the characteristics, its diffusion, economic gains from using it). A subject-based approach is used in the Community Innovation Survey. The 6th CIS contains questions about whether companies are engaged in 10 types of environmental innovation activities but does not provide information on the diffusion of specific innovations, nor does it provide quantitative information about the environmental characteristics. It offers information about the share of companies who have introduced in the last 3 years a product, process, organisational or marketing innovation with environmental benefits in the form of a) Reduced material use per unit of output, b) Reduced energy use per unit of output, c) Reduced CO2 ‘footprint’ (total CO2 production), d) Replaced materials with less polluting or hazardous substitutes, e) Reduced soil, water, noise, or air pollution per output, f) Recycled waste, water, or materials per output, g) reduced energy use for the end-user, h) Reduced air, water, soil or noise pollution for the end-user, i) Improved recycling of product after use. With the help of these questions, the nature and degree of eco-innovation can be studied for 10 environmentally based innovation categories in different Member States. Special permission is needed for using the micro-data. The micro-data are slightly aggregated for reasons of anonymity. Such information helps to establish the prevalence of certain innovations and the role of specific drivers (existing regulation or taxes, anticipated regulation or taxes, demand from customers, government grants, subsidies and other incentives for environmental innovation, “voluntary” environmental agreements). Direct innovation indicators are a highly desirable indicator for macro-environmental assessment by offering information (or an indication) on the prevalence of the innovation, whose environmental and economic characteristics are known by experts or documented in tradejournals and product databases. Information on the economics of use (money being saved or bound) together with information on diffusion can be used to estimate rebound effects. A limitation is that it takes a big effort to collect the data and to aggregate information from different product variants. A direct output measure of eco-innovation can

25


be constructed from announcements in trade journals 11 and product information databases. An example is the green car database established by Yahoo, listing all green car models. The website has a tool called gas mileage impact calculator which allows people to compare cars with one another with respect to fuel consumption and emissions of major pollutants. (see http://autos.yahoo.com/green_center-tech/). A problem for macro-environmental assessment is that few product databases provide information about all relevant environmental aspects (in a standardised way). For specific products, a database of eco-innovation output could be created by sampling the ‘new product announcement’ sections of technical and trade journals or by examining product information provided by producers. Richard Newell followed this approach to determine changes in energy efficiency for air conditioners and gas water heaters Figure 3.

Figure 4. Changes in energy efficiency of new product versions. Source: Newell, Jaffe and Stavins (1999, p. 950).

For vehicles, time series information is available about fuel economy and CO2 emissions per km in standard tests, as a result of the policy interest in these and special reporting requirements. As can be seen from figure 10, CO2 emissions have fallen for gasoline cars and diesel cars. For hybrid-electric cars they increased in the 2004-2006 period.

11

A trade journal or trade magazine is a periodical, magazine or publication printed with the intention of target marketing to a specific industry or type of trade/business. Trade refers to business, not to exports and imports.

26


Figure 5. New passenger cars :CO2 by engine technology. Source: ICCT, 2012, p. 7.

A second type of direct innovation measures, important for calculating macroenvironmental impact, is the diffusion of the innovation. Often good statistics are available at the level of products and technologies. An example is the diffusion of catalytic converters, which is well-documented in terms of penetration (shares of cars with catalytic converter). Diffusion can be measured in three different ways: • Number of products being sold containing the innovation (for instance number of new cars with catalytic converters or number of passive houses being built) • Product sales (in Euro) of an innovative product (gasoline hybrid electric cars) • New generation capacity (Kilowatt) for electric generation technologies (solar PV, ..). For the purposes of macro-environmental assessment, available information from official sources is likely to be aggregate. An example is Figure 4, offering information about CO2 emissions for

all new passenger cars.

27


Figure 6. New passenger cars: CO2 emissions by Member State. Source: ICCT, 2012, p. 3

Although this information certainly can be used to assess trends in environmental impacts related to innovations in the car fleet of EU member states at the macro-level, it is not possible to relate it to any specific innovation at the micro-level. For that, even more disaggregate information is required: per category of car (sports, SUV, large, medium and small sedan, ..) and per fuel type. 2.2.4 Indirect innovation indicators derived from aggregate environmental performance data Eco-innovation can also be measured on the basis of eco-efficiency and resource productivity performance data at higher levels of aggregation than the (functional) product level. Eco-efficiency can be measured on the basis of physical terms (emissions in kilogram for a certain pollutant per km or Kwh) and on the basis of physical and economic terms (when the emissions or waste are considered against an economic value indicator, such as product sales or value added). In the EU, the Eco-Efficiency of Road Transport w.r.t. NOx Emissions, (VKM div by NOx) improved from 1990-2007 (1995=100). Total emissions in CO2 however increased, despite a slight improvement in eco-efficiency. Results based on the Logarithmic Mean Divisia Index (LMDI) (a kind of index decomposition analysis [IDA]) for passenger car CO2 emissions in Europe for gasoline and diesel engines, showed that activity (increase in vehicle km) is the most important factor behind the evolution of CO2 emissions from both gasoline and diesel passenger cars. Technological improvements (regarding energy or carbon intensity) were overwhelmed by emissions stemming from increased activity. The introduction of direct injection helped to reduce CO2 emissions Figure 6.

28


Figure 7. CO2 emissions from diesel cars in the presence and absence of direct injection. Source: EMINInn D6.1

2.2.5. Composite measures Information of the different indicators for eco-innovation may be combined in composite indicators, as is done in the European Innovation Scoreboard, the clean tech database and the eco-innovation survey – the details of which are described below. The European Innovation Scoreboard, provides summary information about the innovation performance of EU27 Member States. Innovation performance in the EIS is measured using data from 29 innovation indicators. These indicators are grouped in three categories: i) Innovation enablers (human resources, quality of research and finance), 2) Firm activities (R&D investments, entrepreneurship and innovation partnerships and intellectual assets), 3) outputs (percentage of innovation SMEs, high growth innovative firms, employment in knowledge-intensive firms, exports of medium and high-tech products, and license and patent revenues as percentage of GDP). The Global Cleantech Scoreboard is an indicator system for measuring drivers and incidence of clean tech at the level of nations. The indicator system is used to generate country profiles. It consists of 4 sets of indicators based on a combination of subjective and objective measures: 1. 2. 3. 4.

General innovation drivers Specific drivers for cleantech innovations Evidence of emerging clean tech innovations Evidence of commercialization of clean tech

29


The GCS 2012 covers half of the twelve new Member States. These new Member States are Poland, Romania, Slovenia, Czech Republic, Bulgaria and Hungary. The Eco-Innovation Observatory offers scores about eco-innovation inputs, activities, outputs, environmental outcomes and socio-economic outcomes for EU MS. An interesting indicator is eco-innovation outputs based on eco-innovation related patents, academic publications related to eco-innovation and coverage of "eco-innovation" in electronic media. Austria is found to have the highest score, followed by the Netherlands Figure 7.

Figure 8. Results from the Eco-innovation scoreboard for Eco-innovation outputs. Source: Eco-Innovation Observatory

The ASEM Eco-Innovation index created by the ASEM SME Eco-Innovation Centre in Korea and offers information on 15 countries. From Europe, Austria, Belgium, Denmark, France, Germany , Italy, Sweden and the United Kingdom are included, togther with China, India, Indonesia, Japan, Korea, Malaysia and Thailand. The ASEM Eco-innovation index is based on 20 indicators: 5 indicators for eco-innovation capacity, 4 indicators for the supportive environment for it, 5 for eco-innovation activities and 6 for ecoinnovation performance. An overview of the indicators is given in Table 3.

30

Environmental Macro Indicators of Innovation Table 3. The ASEM Eco-Innovation Indicators

Source: ASEM Eco-innovation Index 2013 report, p. 36. Results for 2012 are given in Figure 8, showing that innovation capacity, activities and performance are positively correlated. Eco-innovation performance is a composite indicator based on the green industry market size, green jobs in the green technology industry and environmental indicators. For the most part, innovation in indirectly measured.

31


Figure 9. ASEM results for “eco-innovation capacity”, “eco-innovation activities” and and “eco-innovation performance” of analyzed countries. Source: ASEM Eco-innovation Index 2012 report, p. 18

2.2.6 Conclusions and recommendations The macro-environmental impact of an innovation is a function of its use and the characteristics of the innovation (direct effects). An innovation may also have an indirect environmental impact through the demand of related products (such as fuel in the case of cars) and activities that stem from money that is saved or earned through a certain product (income effect). 12 The indirect effects are difficult to estimate with precision but the direct effects are relatively straightforward: impact i is a function of product diffusion, the intensity of use and the environmental characteristics of the innovation. Technology data needed for determining the direct and indirect effects are: • The stock of products (St) where St = St-1 + Nt (Newly bought products) – Dt (Discarded products) or the stock of technologies in the case of producers. • The environmental characteristics of the product or technology at the point of production, use and end-of-lifetime. • Usage of the product (technology). • Usage of related products (e.g. fuels in the case of energy technologies). 12

Energy savings are known to give rise to extra energy use (rebound effect) either through a more intensive use of the energy saving device (direct effect) or by spending the money saved on fuel costs on an overseas holiday. Direct rebound effects are usually fairly small - less than 30% for households. The indirect effects are more difficult to track. When energy efficiency significantly decreases the cost of production of energy intensive goods, rebounds may be considerable. See http://www.ukerc.ac.uk/support/tikiindex.php?page=0710ReboundEffects. A reflection on classes of rebound effects is offered by Turner (2012). In her discussion of the literature she makes the important observation that with regard to ‘economy-wide’ rebound effects “we still lack a rigorous theoretical framework to explain the mechanisms and consequences of the rebound effect at the macro level”.

32


Product lifecycle inventory information can be used to determine the impact at different points of the chain, whereas product databases can be used for information about energy use and emissions at the point of use. Diffusion can be determined from product sales but one should also take into account aspect of use and product exits and to the degree this is possible, interactions with other products (substitution and co-evolution). There is thus a system boundary issue: whether innovation’s performance is studied at the point of use or also at other points of the product life cycle. There is also an demarcation issue for innovation, which can be measured at the level of product components and product families. For researching the macro-environmental impact of innovation top-down decomposition analysis can be used, starting from macro-data, such as energy-efficiency (energy use per unit of output) and bottom-up analysis, starting from product diffusion data and LCA data. An example of a combined approach is the IPAT-LCA decomposition model described in Font Vivanco et al (2014), and also described further in section 5.

3.3 Innovation and environmental pressures: a simple model An early internal background paper for EMInInn set out a simple model for considering how incentives and innovation system conditions drive innovation outcomes, with consequent environmental pressures (see Box 1 for categorisation of environmental indicators). Box 1. DPSIR Human interventions into the natural environment can be measured and characterised in a wide variety of ways. In order to simplify this complexity, the EEA has adopted the a framework for categorising indicators of such interventions. • • • • •

Driving forces (e.g. vehicle kilometres travelled) Pressures (e.g. gCO2 emitted) State (e.g. atmospheric concentration of GHGs) Impact (increased global temperatures) Response (increased expenditure on hurricane damages)

For several pragmatic reasons, it was determined within EMInInn to focus on pressure indicators, since these are the most straightforward common basis on which to compare innovations. The model, developed in analogy to the DPSIR framework of environmental interventions, is shown in Figure 9. It was used as a starting point to think about the factors that shape the direction and rate of innovation, and the environmental pressures and societal responses that occur as a result of the diffusion of innovations. 33


Actors, involved in systems of innovation (which comprise actors and technologies, the networks through which they interact, and the institutions that govern those interactions), respond to incentives and opportunities. Their activities lead to a variety of complex innovation processes, including the generation and dissemination of new knowledge; the invention of new products, processes, institutions, marketing methods, etc.; the introduction of those novelties into the economic realm (i.e. innovation), and the diffusion of those novelties into society. As technologies and practices diffuse, they both undergo further change as innovation system actors respond to new knowledge concerning the nature of demand, the real-world performance of products, etc. In other words, the innovation process is iterative and non-linear, involving multiple feedbacks within the innovation system. Moreover, as the new good/service diffuses, the production and consumption of the new goods/services results in environmental pressures, which are likely to differ from the environmental pressures generated by the production and consumption of a relevant alternative. This influence can be a net positive or a net negative one and how to discriminate between the two is a core aim of EMInInn. At the same time, the diffusion of the good/service will result in a series of economic feedbacks, through changes mediated by consumers (who may be paying more or less for the new good/service, with resulting shifts in their spending and saving behaviour) and other producers (who may face higher or lower prices arising from changes in the demand for inputs required to produce the new good/service). These economic responses also generate environmental pressures, which can be net negative or net positive.

34


Incentives and barriers: market opportunities, technical possibilities, institutions, etc. Incentives lead to innovation system responses Changing incentives for innovation

Innovation processes Invention

Market changes

Economic feedbacks

Diffusion & Adoption Changing Direct pressures

Changing Indirect pressures

Policy influences incentives and conditions Policy influences innovation processes

Policy Responses Policymakers monitor environmental pressures

Environmental pressures Figure 10. Simple model for considering the relationship between innovation and environmental pressures

Finally, there are additional feedbacks that change innovation processes themselves. These derive from both the innovation processes themselves (e.g. learning by doing improves the prospects fo the technology, thus increasing incentives to invest in it) and from the economic feedbacks associated with diffusion (e.g. all else being equal, one might expect a significant uptake of electric vehicles to generate a fall in petrol price, with a resulting change in the incentives for innovation). There are also feedbacks generated through policy, since policymakers act to change innovation incentives, and they do so in part in response to monitoring both environmental pressures and innovation processes themselves. These feedback responses can also directly influence the innovation process, e.g. new technological provisions embodied in policies as new emission limits, new quantified targets, etc. This model, although a simplification, provides a depiction of the multiple feedbacks and processes that determine the relationship between innovation processes and changes in environmental pressures. The key issue, from the perspective of EMInInn, is to identify the methods and indicators by which these feedbacks and processes can be monitored and assessed. Several key points can be drawn from the model. 35


•

•

•

•

•

Innovation is not a single, linear process, but the product of a complex system of innovation in which many processes occur. Many innovation indicators in widespread use (R&D spending, patents, publications, proportion of revenue attributable to new products, etc) capture just some elements of these processes, including some insight into their direction and rate. However, the complexity and diversity of innovation processes has meant that no single indicator is appropriate for measuring innovation as a whole, with the result that there are a proliferation of ‘scoreboards’ and sets of indicators that seek to capture various dimensions of innovation processes. The changes in environmental pressures that arise from the diffusion of innovations are a result not only of the direct characteristics of the innovation itself (and its relevant alternatives), but also of the indirect effects, mediated by economic feedbacks, which are themselves influenced by policy frameworks. Methods used to understand the environmental pressures associated with the diffusion of innovations vary in the extent to which they are able to take into account economic feedback processes, such as ‘rebound effects’. From a policy perspective, it is clearly important to have clarity about which relationships have been accounted for, and which are left outside of any formal analytic approach. Policy action in response to changes to environmental pressure may result in changes to the economic feedbacks that occur as an innovation diffuses into society. It therefore seems desirable that assessments of the expected economic feedbacks are explicit about the wider policy framework that has been assumed. Policymakers also monitor both innovation processes and environmental pressures, and change policies in response to observed outcomes. Where indicators used in monitoring fail to provide a complete picture of the relevant processes linking innovation to environmental performance, it is likely that they will be misleading 13. Some existing indicators take for granted certain relationships in the framework and ignore others. For example, indicators that focus solely on green patents, and assume that these represent ‘green innovation’, assume that there is a straightforward relationship between patenting, the introduction of new products and processes, and the diffusion of those such that environmental pressures are reduced. At the same time, looking at specific innovations (‘technology assessment’) that are resourcesaving according their technical features (‘green technologies’) without taking into account the complexities of (changing) pressures associated to their large-scale diffusion in the economy raises the risk to give policy support to innovations that have net negative pressures when fully deployed at the system level.

13

In EMInInn, for road transport, we examined the correlation between patents and the eco-efficiency measures. Whereas the results of the correlation analysis showed a strong correlation between patents and eco-efficiency, in the panel regressions for vkm/CO2eq and vkm/NOx, the number of Triadic Patents in emissions abatement and fuel efficiency in transportation is not significant in any of the models (for details of the analysis and details about the results see D2.3). In the panel regression results for vkm/NOx, the share of catalytic converters in the car fleet again is the most significant variable in the analysis. The positive sign is also expected since the catalytic converters reduce nitrogen oxides by about 76.5%. The emissions of NOx from diesel vehicles are 32.6% higher than for gasoline cars and this is also reflected in the significant negative effect of diesel vehicles on the eco-efficiency of road vehicles with respect to NOx at 7% level. The Number of Triadic Patents in Emissions abatement and fuel efficiency in transportation is not significant. The positive partial correlation found between patents and eco-efficiency can be said not to reflect a causal relationship.

36


The EMInInn project has applied existing tools, and developed new tools, to examine a variety of the relationships and interactions depicted in the simple framework, and to identify potential indicators for monitoring these processes and assessing their likely past or future development.

3.4 The EMInInn approach and analytical framework for case studies 3.4.1 The approach Policymakers interested in the relationship between environment and innovation are typically concerned with designing policy interventions that facilitate innovation that generates environmental benefits. Understanding the relationships between the processes in the simple model shown in Figure 9 and monitoring their outcomes, is important for designing such policies effectively. EMInInn did not aim at producing one single model to support this policymaking aim. Instead, a variety of tools and indicators applied in policy analysis and developed or applied within EMInInn help to inform the relationships between the elements in Figure 9, and the way in which they can be assessed and monitored. Tools and indicators for informing policymakers about the environmental outcomes associated with innovation can largely be distinguished between the following 14: 1. Assessment and monitoring of innovation processes in terms of related environmental implications. This kind of innovation assessment and monitoring helps policymakers to identify whether the innovation system and wider conditions at the macro-level are delivering ecoinnovation, without identifying specific innovations. This kind of assessment examines the relationships and processes illustrated in Figure 9, and enables understanding of the conditions (and policies) under which innovation processes, including the diffusion phase, lead to reductions in environmental pressures; 2. Assessment of specific innovations. This kind of technology assessment is widely used to inform the design of policies and R&D priorities, but EMInInn added a specific focus on diffusion phases in order to capture indirect environmental effects and feedbacks, e.g. rebound, also for some specific technologies. Each mode of assessment and monitoring can inform different stages of the policy cycle, and each addresses different kinds of policy relevant question pertaining to measurement and assessment of the environmental pressures associated with innovation.

14

These were described as “viewpoint 1” and “viewpoint 2” within the project, see EMInInn background paper 2 for a discussion.

37


Table 4 illustrates the kinds of questions that might arise across the policy cycle and under the different modes of assessment. Most of the analytic work within the EMInInn project focused on methods and tools that support impact assessment and evaluation of specific technologies; with some work (notably within WP8) focused on tools and methods that support monitoring and evaluation of the environmental consequences of innovation processes, and that attempt to understand the conditions under which innovation processes yield macro-environmental improvements. However, an ex post systemic perspective at the meso scale, which includes pressure changes associated to economic feedbacks (rebound), has been developed in the analytical workpackage on transports (WP6). An even broader system level analysis has been developed, in an ex ante perspective, in WP9 on modelling, in which diffusion scenarios have been embodied in a macro-level CGE model.

38

Environmental Macro Indicators of Innovation Table 4. Illustrative policy questions and methods for the two basic sets of questions addressed within EMInInn

IA (ex ante)

Evaluation (ex post)

Analysis of specific innovations Analysis of innovation process (Technology Assessment) Questions: Questions: - What are the expected environmental - What are the expected innovation pressures from deployment of this effects of environmental policy A vs. technology/this technology-specific policy B? policy? - What are the expected environmental benefits of innovation policy X? Methods: - Scenario-based LCA, EE-CGE, sectorMethods: specific models, etc. - Modelling induced innovation, using econometrically-estimated factors. Questions: - What were the environmental pressures that arose as a result of this technology/technology-specific policy? Methods: - Same tools as ex ante analysis but applied ex post - Decomposition analysis using technology-specific data

Monitoring Questions: - How is the technology working in practice (e.g. biofuel sustainability criteria; RHI heat monitoring data) Methods: - Ad hoc studies, reporting requirements

Questions: - Has environmental policy A induced innovation? - Has innovation policy X contributed to reductions in environmental pressures? Methods: - Identifying environmental outcomes of innovation policies: - Identifying induced innovation: Analysis of patents; analysis of innovation system responses; case studies; Questions: - is innovation contributing to a net decrease in environmental pressures - Is innovation moving towards ecoinnovation? Methods: - Decomposition; - Econometrics & statistics

3.4.2 Guidance for analytic case studies: the EMInInn analytic framework EMInInn developed a common analytical framework to be use in addressing the kinds of questions outlined above through the case studies in WP4-8 (energy, Internet, transport, building sector, waste). Case studies of EMInInn conformed to this analytical framework while taking into account the specific features of the case studies and then exploiting the analytical framework in a flexible way. The analytical framework is made of reasrch questions, criteria, and guidelines for analysis,, which can be summarised by the following elements: Research questions 39


Two categories of research questions, as corresponding todifferent “viewpoints”, are distinguished and used as reference questions on case studies: •

•

Questions relating to innovation as a process: Under what conditions will innovation lead to environmental improvements, compared to an appropriate reference situation without the innovation, including conditions for adoption and diffusion of the innovation? Questions relating to specific innovations. How can the environmental impacts from specific, concrete innovations be assessed, compared to an appropriate reference situation without the innovation(s), including observed diffusion and application?

Boundaries of the system Two types of systems are distinguished: regional and functional systems. A regional system is defined by geographical boundaries (i.e. countries, country groups, the EU) and includes all processes and flows within that region within a defined time period. A functional system is defined by a specific function and includes the cradle-to-grave chain connected with that function. A reference situation is defined: the environmental performance of the identified system with innovation is compared to an alternative system without the innovation in such a way that comparison is possible and meaningful. For both systems environmental pressure are provided . For regional systems, the temporal and spatial definitions of the system with innovation and the reference system are the same. For functional systems, the function is the same in the system with innovation and the reference system. Environmental indicator(s) The environmental indicators are defined at the level of Pressures in the DPSIR chain: emissions of pollutants, extractions of resources and land use. The environmental indicators represent an important pressure exerted by the system, which is affected by the implementation of a new or significantly improved product (good or service), a new process, a new marketing method, or a new organizational method., or the widespread diffusion of an existing technology that is new for the adopters;. Innovation indicator(s) Three types of innovation indicators are distinguished. i.

ii.

Input measures of innovation: Research and development (R&D) expenditures, R&D personnel, and innovation expenditures, all relating to the implementation of a new or significantly improved product (good or service), or process, a new marketing method, or a new organizational method. Intermediate output measures of innovation: the number of patents; numbers and types of scientific publications, all relating to the implementation of a new or significantly improved product (good or service), or process, a new marketing method, or a new organizational method.

40


iii.

Direct output measures of innovation: data on sales or stock of new products (adoption, diffusion), relating to the implementation of a new or significantly improved product (good or service), or process, a new marketing method, or a new organizational method.

The analytical framework prescribes that for both the system with the innovation and the reference system without it (counterfactual), measured, calculated or estimated indicators of environmental pressure and innovation are generated in such a way that comparison is possible and meaningful for the research questions. The methods applied and the data used for such ana analysis must be clearly documented and should be chosen appropriate to the research questions and the defined system with reference to the state fo information on the chosen case study.. Given the two categories of research questions, the interpretation of the result will fall in one of two categories: i.

Under what conditions will the implementation of new or significantly products (good or service), or processes, new marketing methods or a new organizational method - as measured by input measures of innovation, intermediate output measures of innovation and/or direct output measures of innovation and through diffusion - lead to a decrease in emissions of pollutants, a decrease in extractions of resources and/or a decrease in land use at the macro-level? In short: under what conditions innovation leads to environmental improvements at the macro-level? Did the implementation of a new or significantly product (good or service), or process, a new marketing method of a new organizational method - as measured in terms of diffusion - lead to a decrease in emissions of pollutants, a decrease in extractions of resources and/or a decrease in land use at the macro-level? In short: How can the environmental pressures of innovations be assessed at the macro-level?

ii.

The interpretation under (i) is bout the 'conditions' for observing decreasing pressures from innovation, which can be related to the factors driving the invention/diffusion of innovations, and how these factors should work in a way able to decrease environmental pressures substantially and quickly, also in relations to policy objectives and targets. The interpretation under (ii) is about assessment methods and measurability in oder ato achieve reliable results on the main question of net environmental pressure change from innovation The expected major results are of four types, all of them representing advances compared to the present state of knowledge: I.

II.

Net macro-level environmental pressure from innovations: Given an innovation, the results should show the macro-level changes in environmental pressure as a result of its application in society, directly, indirectly and via rebound and other feedeback mechanisms. Innovation and macro-efficiency indicators: The results show the contribution of certain innovation(s) to observed changes in environmental pressure at the macro level, as measured for example by resource efficiency indicators at the macro-level (e.g. GHG emissions on GDP, DMC on GDP). 41


III.

IV.

Joint economic and environmental innovativeness indicators: The results show a (causal or statistical) link between certain societal conditions and policies, the occurrence of innovations in that society and the fraction of innovations that can be reliably defined as eco-innovations (net drecrease of environmental and resource pressures). In addition, the results can also provide Inputs for better macro ex ante modelling of environmental and economic dynamics (e.g. CGE models) in order to improve scenario making.

42


4 Current policy practice This section outlines current policy practices on EU level regarding environmental assessment at the various stages of the policy cycle. Having an overview of methods and indicators currently used and being aware of shortcomings of the current system can allow identifying areas where the contributions of EMInInn could be most useful. To ensure that EU action is effective the EU has established within its Smart Regulation framework a number of requirements relating to the development, monitoring and evaluation of its actions. Policies, EU laws, and other measures are usually assessed during the following stages of the policy cycle (see Figure 3): • • •

During the policy formulation stage through an ex ante Impact Assessment (IA) During the implementation stage through Monitoring During the evaluation stage through an Ex-post Evaluation

Figure 1 Assessment of environmental effects throughout the policy cycle

Agenda setting & problem identification

Evaluation

Policy Formulation

Implementation

Decisionmaking

For each of these three assessment moments different requirements apply. The following section outlines the existing requirements and guidance on EU-level relating to the assessment of the environmental effects of innovation, and sketch current practice in their application. 15 It should be noted that this chapter is limited to the technical requirements concerning the abovementioned assessment tools. Examining the political forces, such as political bargaining, that 15

Please note that Environmental Impact Assessments (EIAs) and Strategic Environmental Assessment (SEAs) will not be examined here as they do not apply to EU-level interventions but are used on the Member State level.

43


eventually determine the final decision-making falls outside of the scope of EMInInn. Therefore readers should keep in mind that recommendations stemming from the tools examined here (IAs, monitoring, ex post evaluations) are only one factor influencing decision-making. For example even if the Impact Assessment of a certain policy options is very positive, this does not necessarily mean that this option will go through (see section 4.1.3.1 for a recent example). While politicians usually take assessments into account, they might come to a very different conclusions; for reasons that are studied in the realm of political science and will not be examined here. So even though the tools examined in this chapter do not always drive decision-making, they clearly often have a role to play. Hence it is important to examine existing requirements and practices as they are the primary point of reference for EMInInn to inform and improve current policy practice.

4.1 Impact Assessments Impact assessments (IA) are procedures used to analyse potential effects of new policies before their adoption. After being provided for by the Treaty of the Union (Art. 174) 16, IAs have been introduced in 2002 for the Commission proposals. In practice, they have received more attention than other regulatory management tools, such as ex post evaluation. 4.1.1

Requirements

The requirements relating to impact assessments are laid down in the Commission’s Impact Assessment Guidelines 17 (2009). In principle, not all EU policy proposals involve impact assessments. In general, IAs are necessary for the most important Commission initiatives and those which will have the most far-reaching impacts. For which initiatives an IA needs to be carried out is decided each year by the Secretariat General/Impact Assessment Board and the departments concerned. Concerning the actual assessment of impacts the guidelines outline the following steps: • • •

Step 1: Identification of economic, social and environmental impacts Step 2: Qualitative assessment of the more significant impacts Step 3: In-depth qualitative and quantitative analysis of the most significant impacts

Policy makers are advised to screen identified policy option for a range of environmental impacts next to economic and social impacts (see Table 5). Table 5 Overview of environmental impacts to be considered as part of the Impact Assessment Variable

Suggested question to consider in IA of interest to EMInInn

• Economic impact Innovation and • Does the option promote greater productivity/ resource efficiency? research • Environmental impact

16

“In preparing its policy on the environment, the Community shall take account of: available scientific and technical data; […]; the potential benefits and costs of action or lack of action […]”.

17

http://ec.europa.eu/smart-regulation/impact/commission_guidelines/docs/iag_2009_en.pdf

44


The climate

Transport and the use of energy

• • • • • • •

Air quality

Biodiversity, flora, fauna and landscapes

• •

• • •

Water quality and resources

• • •

Soil quality or resources

• • •

Land Use

• •

Renewable or non-renewable resources

•

The environmental consequences of firms and consumers

• •

•

• •

Waste production/

•

Does the option affect the emission of greenhouse gases (e.g. carbon dioxide, methane, etc.) in to the atmosphere? Does the option affect the emission of ozone-depleting substances (CFCs, HCFCs)? Does the option affect our ability to adapt to climate change? Will the option increase/decrease energy and fuel needs/consumption? Does the option affect the energy intensity of the economy? Does the option affect the fuel mix (between coal, gas, nuclear, renewables, etc.) used in energy production) Will it increase or decrease the demand for transport (passenger or freight), or influence its modal split? Does it increase or decrease vehicle emissions? Does the option have an effect on emissions of acidifying, eutrophying, photochemical or harmful air pollutants that might affect human health, damage crops or buildings or lead to deterioration in the environment (soil or rivers etc.)? Does the option reduce the number of species/varieties/races in any area (i.e. reduce biological diversity) or increase the range of species (e.g. by promoting conservation)? Does it affect protected or endangered species or their habitats/ ecologically sensitive areas? Does it split the landscape into smaller areas or in other ways affect migration routes, ecological corridors or buffer zones? Does the option affect the scenic value of protected landscape? Does the option decrease or increase the quality or quantity of freshwater and groundwater? Does it raise or lower the quality of waters in coastal and marine areas (e.g. through discharges of sewage, nutrients, oil, heavy metals, and other pollutants)? Does it affect drinking water resources? Does the option affect the acidification, contamination or salinity of soil, and soil erosion rates? Does it lead to loss of available soil (e.g. through building or construction works) or increase the amount of usable soil (e.g. through land decontamination)? Does the option have the effect of bringing new areas of land into use for the first time? Does it affect land designated as sensitive for ecological reasons? Does it lead to a change in land use (for example, the divide between rural and urban, or change in type of agriculture)? Does the option affect the use of renewable resources (fish etc.) and lead to their use being faster than they can regenerate? Does it reduce or increase use of non-renewable resources (groundwater, minerals etc.)? Does the option lead to more sustainable production and consumption? Does the option change the relative prices of environmental friendly and unfriendly products? Does the option promote or restrict environmentally un/friendly goods and services through changes in the rules on capital investments, loans, insurance services etc.? Will it lead to businesses becoming more or less polluting through changes in the way in which they operate? Does the option affect waste production (solid, urban, agricultural, industrial, mining, radioactive or toxic waste) or how waste is treated, disposed of or recycled?

45


generation/ recycling The likelihood or scale of environmental risks Animal welfare

International environ-mental impacts

• • • • • •

Does the option affect the likelihood or prevention of fire, explosions, breakdowns, accidents and accidental emissions? Does it affect the risk of unauthorised or unintentional dissemination of environmentally alien or genetically modified organisms? Does the option have an impact on health of animals? Does the option affect animal welfare (i.e. humane treatment of animals)? Does the option affect the safety of food and feed? Does the option have an impact on the environment in third countries that would be relevant for overarching EU policies, such as development policy?

Source: extracted from European Commission (2009), Impact Assessment Guidelines

An important point to highlight is that according to the IA guidelines the responsibility to establish the proportionate level of analysis (depth, scope, resources used) lies with the author of the IA. He or she is free to choose what is in his/her eyes the most appropriate level of analysis, based on past experience, available data, external expertise, and preliminary contacts with stakeholders. 18 Quality control of IAs (including the assessment of environmental impacts) is carried out by the Impact Assessment Board. The Board examines and issues opinions on all draft impact assessments. In principle, a positive opinion is needed from the Board for an initiative to be tabled for adoption by the Commission. All impact assessments and all IAB opinions are published once the Commission has adopted the relevant proposal. 4.1.2 Guidance The Impact Assessment Guidelines are supported by a number of guidance document that aim to support policy officials and external evaluators in carrying out IAs, such as: •

• •

Annex 1-13 of the Impact Assessment Guidelines 19 (2009): Annex 9 provides more detailed information on how to assess non-market impacts in particular on the environment and health. Information is provided on topics such as how to monetize the cost of carbon emissions (p. 44) and the use of Life Cycle Assessment (p. 45). Best Practice Library (part of Annex 14 of the Impact Assessment Guidelines): This regularly updated library provides examples of Impact Assessments currently identified as best practice. IA tools website: This online platform provides more in-depth information along four modules: 1) Impact Inventory (provides links to data sources for the different impact areas); 2) Model Inventory (guidance on which models to use when assessing a certain impacts); 3) Good Practice Inventory; 4) IA tool handbook. While the platform was made public in 2008 20 it does not appear to be available any longer.

18

http://ec.europa.eu/smart-regulation/impact/commission_guidelines/docs/iag_2009_en.pdf http://ec.europa.eu/smart-regulation/impact/commission_guidelines/docs/ia_guidelines_annexes_en.pdf 20 https://ec.europa.eu/jrc/sites/default/files/jrc_newsrelease_20080423_iatools.pdf 19

46


It is striking that while there is a separate and very detailed guidance document on the assessment of social impacts 21, a similar document concerning environmental impacts appears not to be available on the Commission’s website. In 2013 Erawatch published “A guidebook to assessing environmental impacts of research and innovation policy” 22 which was prepared by Technopolis Group. The 274-page long document provides detailed guidance on how to assess environmental impacts (ex ante and ex post) on programme, instrument, and project level. Moreover, several environmental assessment methods such as Life Cycle Assessment (LCA) are described. It appears surprising that this guidebook is not mentioned on the Commission’s website, as it could be of tremendous help for policy makers to assess environmental effects of innovation. In any case it seem advisable to promote this guidebook more widely so that it is well-known in the policy evaluation community. 4.1.3 Current practice In general it seems that the requirement to assess environmental impacts as set out in the Commission’s IA Guidelines and the subsequent quality control of all IAs by the Impact Assessment Board provides a sufficiently strong incentive for policy makers to consider the environmental effects of planned initiatives. However, in practice environmental assessments are often a “tick-thebox” exercise and potential costs and benefits are not quantified. This might be attributed to the lack of methods for quantification or monetarisation that policy makers perceived in the past. However, positive examples of thorough environmental assessments, including forms of CBA, exists, such as the IA for the Strategic Energy Technology (SET) – Plan and the IA for the new proposed Waste Directive associated to the circular economy package (2014) 23. Also the “Guidebook to assessing environmental impacts of research and innovation policy” 24, which was published by Erawatch in 2013, might counter this perceived lack of methods. The following section will outline policy current practices more in detail. A 2007 evaluation 25 of the Commission’s IA system points to methodological difficulties policy makers experience in assessing environmental impacts. A survey of Commission officials and stakeholders show that 11% of Commission officials agreed, while 38% disagreed, that appropriate tools (i.e. methodologies, analytical models, support groups within the Commission etc.) are in place to assess the environmental impacts of proposals. Commission officials stressed that there is “general lack of tools and methodologies to adequately quantify or even monetise environmental benefits. Even where tools and methods exist, they tend to be very costly and time-intensive to apply, and can give rise to differences of opinion (e.g. regarding how certain environmental impacts should be expressed in monetary terms).” Around 40% of the surveyed stakeholders felt that “the IA system’s approach should be refined to facilitate a more in-depth analysis of the environmental impacts (e.g. impacts on air, water and soil quality; biodiversity; the climate; animal and plant health; etc.).”

21

http://ec.europa.eu/smart-regulation/impact/key_docs/docs/guidance_for_assessing_social_impacts.pdf http://europa.eu/!Xm84ND; summary: http://ec.europa.eu/research/evaluations/pdf/archive/other_reports_studies_and_documents/envti0413167enn_002.pdf 23 http://ec.europa.eu/environment/circular-economy/ 24 http://europa.eu/!Xm84ND; summary: http://ec.europa.eu/research/evaluations/pdf/archive/other_reports_studies_and_documents/envti0413167enn_002.pdf 25 http://ec.europa.eu/smart-regulation/impact/key_docs/docs/tep_eias_final_report.pdf 22

47


Also the annual reports of the Impact Assessment Board (IAB) indicate that environmental assessments are an issues. The 2013 report shows that the number of comments of the IAB relating to the assessment of environmental impacts has risen from 10% in 2010 to around 30% in 2012 and 2013. This might indicate that the quality of environmental assessments has decreased, or that the IAB is increasingly critical concerning the assessment of environmental impacts. It seems more likely that the latter is the case. Also a review of a number of milestones produced through EMInInn (MS 18, 19, 20, 21) covering a range of EU policy areas (innovation policy, energy and climate policy, biodiversity policy, resource and waste) and a review of several individual IAs by the authors highlights several shortcomings in the current system. The most important results of the analysis are: • There is a variety of methods used to assess environmental effects. They range from modelling (e.g. PRIMES model for the SET-plan), to simple assessments based on assumptions and expert judgments (or background studies) on the changes of relevant parameters. It is worth noting that methods are evolving in some cases. For example the IA of the 'Thematic Strategy on the prevention and recycling of waste' (EC 2004) included just a ‘simplified model’ to assess, under certain assumptions, the GHG effects of increasing recycling/recovery, whereas the proposed Waste Directive of 2014 exploit the first results of a recently produced ‘European Reference Model on Municipal Waste Management’ 26. • Until recently environmental effects did not appear to receive specific attention in innovation policy making (with the exception of eco-innovation). For example the Handbook for ‘SMART Innovation‘ (2006) does not include specific provisions or metrics for the evaluation of the environmental impacts of innovation programmes. However, the recent publication of “A guidebook to assessing environmental impacts of research and innovation policy” 27 by Erawatch seems to close this gap. Though it remains to be seen how actively it will be promoted and used by the Commission. • Most of the reviewed impact assessments of initiatives in the realm of innovation policy seem to mention positive environmental impacts, and not the negative ones. In practice, it is often the case that environmental impacts are neglected under the assumption that 'the initiative is environmentally friendly, and no negative environmental impacts are expected'. For example the IA for the European Research Area provides only a brief qualitative descriptions of environmental impacts, without any real quantification. • Some examples of specific ex ante assessment of expected environmental impacts associated with R&I policies do exist, such as the estimated emissions reductions associated with the SET-Plan and the proposed Waste Directive for a ‘circular economy’, but these appear to be the exception. • Indirect environmental effects (e.g. rebound effects) are rarely assessed. This bears the risk that initiatives are judged more positively from an environmental perspective than they might deserve. • Many IAs also do not take account of the fact that there might be trade-offs between environmental pressures. For example a positive effects in one environmental area might lead to negative effects in another area. Ignoring such trade-offs can have serious environmental consequences, as illustrated in the case of the EU’s biofuels policy where the 26 27

http://www.wastemodel.eu/

http://europa.eu/!Xm84ND; summary: http://ec.europa.eu/research/evaluations/pdf/archive/other_reports_studies_and_documents/envti0413167enn_002.pdf

48


•

IA concerning the 10% biofuels target didn’t consider the impact of increased biofuel production on land-use (see Box 1). A positive example of how evidence on trade-offs stemming from IAs can influence final decision making comes from the UK. During the IA of the Renewable Heat Incentive 28 it became apparent that increased biomass combustion would have significant negative effects on air quality and human health. As a consequence the policy was revised to include an emissions standard for biomass heating technologies which would reduce the emission of air pollutants. In the past Impact Assessments often did not specify environmental targets, nor did they establish monitoring and evaluation frameworks. (MS 21, p. 27)

Box 1: Failing to account for trade-offs between environmental pressures – the case of the EU’s biofuels policy To address climate change the EU established a 10% in 2020 target for biofuels in its 2006 Renewable Energy Roadmap. The impact assessment for the Roadmap included analysis from a wide variety of models, including PRIMES, POLES-PACE, ASTRA, a dedicated renewable energy model called Green-X, and a dedicated input-output model used to estimate employment effects. All of these assumed land-use to be constant and exogenous, and thus none were able to directly examine the potential for the policy to influence land-use change indirectly via price changes, despite these price changes being modelled extensively for the purposes of exploring GDP and employment effects. In short, despite the use of a wide array of analytic tools, the analysis failed to explore the potential impacts of the policy on the environmental performance of intimately related sectors. The assessment of the expected biodiversity impact of the policy neglected this potential effect on land, and yet was emphatic in its declaration that “it is certain that the effect of the [policy] is substantially positive”. In retrospect, such a conclusion was not supported by adequate analysis, and greatly overstated the degree of confidence that was justified by the available analytic tools. This failure to consider potential effects on land-use has since then attracted much criticisms of environmental activists but also the mainstream media. It is argued that farmers (mostly in developing and middle income countries) are incentivised to expand agricultural land by hacking into rainforests and draining wetlands. 29 To address these concerns the EC has commissioned several studies that explicitly address land use changes, such as “Global Trade and Environmental Impact Study of the EU Biofuels Mandate” (2010) and “Assessing the Land Use Change Consequences of European Biofuel Policies” 30 (2011). The latter study finds that the biofuels policy leads to an increase in cropland area between 1,73 and 1,87 million hectares (depending on trade liberalization). The most affected region is Latin America (primarily Brazil). The study also shows that 20% of the change in land use can be attributed to the exploitation of new land such as savannah, grasslands and primary forest. The most striking result is that emissions resulting from changes in land-use eliminate more than two-thirds of the direct emission savings generated by the policy. The biofuels case highlights that failing to consider trade-offs between environmental pressures might have serious environmental implications that can reduce net environmental benefits to a large extent. Therefore such trade-offs should be considerate in Impact Assessments, preferably through incorporating them in modelling. Moreover, IAs should better document which indirect effects are 28

https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/48042/1381-renewableheat-incentive-ia.pdf 29 http://www.euractiv.com/sustainability/controversy-mounts-eu-over-fall-out-biofuel 30 http://www.corbey.nl/files/media_base/original/103.pdf

49


being taken into account, which can be envisaged but are excluded from formal analysis, and provide some indication of the sensitivity of findings to key uncertainties. 4.1.3.1 Example 1: Impact Assessment of the Circular Economy Package Background In February 2014 the European Commission issued a Communication entitled ‘Towards a circular economy: a zero waste programme for Europe’ 31. Its stated aims were “to establish a common and coherent EU framework to promote the circular economy”, thereby: • • • •

boosting recycling and preventing the loss of valuable materials; creating jobs and economic growth; showing how new business models, eco-design and industrial symbiosis can move us towards zero-waste; reducing greenhouse emissions and environmental impacts.

As part of what came to be called ‘the circular economy package’, the Commission also put forward a legislative proposal to satisfy its requirement to review recycling and other waste-related targets enshrined in a number of waste-related Directives: the Waste Framework Directive 2008/98/EC, the Landfill Directive 1999//31/EC and the Packaging and Packaging Waste Directive 94/62/EC. The main elements of the proposal were to: • • • • • • • • • •

Increase recycling/re-use of municipal waste to 70% in 2030; Increase packaging waste recycling/re-use to 80% in 2030 with material-specific targets set to gradually increase between 2020 and 2030; Phase out landfilling by 2025 for recyclable waste corresponding to a maximum landfilling rate of 25%, with a reduction to 5% landfilling of residual waste by 2030; Reduce food waste generation by 30% by 2025; Introduce an early warning system to anticipate and avoid possible compliance difficulties; Ensure full traceability of hazardous waste; Increase the cost-effectiveness of Extended Producer Responsibility schemes by defining minimum conditions; Simplify the reporting obligations and lighten obligations affecting SMEs; Harmonise and streamline the calculation of the targets and improve the reliability of key statistics; Improve the overall coherence of waste legislation by aligning definitions and removing obsolete legal requirements.

In the event, the ‘circular economy package’ was withdrawn by the incoming Commission in December 2014, in order to permit the adoption of “a broader and more ambitious approach that can be more effective”. This step has created much commotion among environmental NGOs, but also in the European parliament and among national governments. NGOs suspect that the Commission’s decision to withdraw that package was influenced by the lobby organisation

31

http://ec.europa.eu/environment/circular-economy/. The Impact Assessment of the proposals in the Communication can

also be found here.

50


BusinessEurope which suggests in a letter to the Commission that the circular economy package “should be withdrawn and re-tabled as an economic piece of legislation.” 32 Many critics also fear that the new Commission’s focus on growth and jobs will lead to the subordination of environmental concerns. 33 However, shifting the focus to economic targets appears not to be a valid argument to withdraw the package as the Impact Assessment concluded that achieving the proposed new waste targets “would create 180 000 new jobs, while making Europe more competitive and reducing demand for costly scarce resources”. Others think that with the withdrawal the Commission aims to address concerns of European citizen that the EU extending is powers and that it should refocus on its core business; the internal market. In summary, the scrapping of the circular economy package illustrates that the policy recommendations provided in Impact Assessments are not always adopted, but that many other factors (e.g. lobbying, changing political priorities, public opinion) influence political decision-making. The Impact Assessment and its methodology is now discussed in more detail. Impact Assessment The conclusions of the Impact Assessment of the circular economy package were that it would lead to: •

Reduction in the administrative burden, in particular for SMEs, through simplification and better implementation including by keeping the targets ‘fit for purpose’

•

The creation of more than 180.000 direct jobs by 2030

•

GHG emission reduction of around 443 millions of tons of GHG between 2014 and 2030

•

Positive effects on the competitiveness of the EU waste management and recycling sectors as well as on the EU manufacturing sector

•

Marine litter levels 13% lower by 2020 and 27.5% lower by 2030

•

Reinjection into the EU economy of some 50 million tonnes more secondary raw materials (paper/cardboard, plastics, metals, glass) than were recycled in 2011.

The various options assessed were as follows: Option 1: Ensuring full implementation of existing EU target commitments Option 2: Simplification, improved monitoring, diffusion of best practices Option 3: Upgrading EU targets, with various sub-options: Option 3.1: Increase the recycling/reuse target for municipal waste by 2030 to 70% (or 60%) Option 3.2: Increase the re-use/recycling targets for packaging waste by 2030 to 80% overall reuse/recycling) Option 3.3: Phasing out landfilling of recoverable municipal waste (ban on plastic/paper/glass/metals by 2025 (max 25% landfilling), global ban by 2030 (max 5%)) Option 3.4: Combination of options 3.1, 3.2 and 3.3 Option 3.5: Same as option 3.4 with different deadlines for different groups of countries Option 3.6: Same as option 3.4 with more stringent deadline for all MS with the possibility of time derogation for some MS Option 3.7: Same as option 3.4 with landfill ban on all similar waste 32

http://www.euractiv.com/files/businesseurope_statement_to_the_new_commission__business_input_to_the_screening_exercise_by_vice-president_timmermans.pdf 33 https://qceablog.wordpress.com/2014/12/19/the-circular-economy-and-better-regulation/

51


The Impact Assessment was carried out using a specially constructed model. The overall schematic of the model is given in Figure 10. The model is to be a permanent tool, maintained and improved by the European Environment Agency.

Figure 11. Overall schematic review of the model used in the waste policy review for the circular economy package. 34 Source: Gibbs et al 2014, Figure E1, p. ii

Figure 11. shows the net social costs of each option compared to Option 1, the full implementation scenario. 34

Adrian Gibbs (Eunomia), Tim Elliott (Eunomia), Thomas Vergunst (Eunomia), Ann Ballinger (Eunomia), Dominic Hogg (Eunomia), Emmanuel Gentil (CRI), Christian Fischer (CRI), Ioannis Bakas (CRI) 2014 ‘Development of a Modelling Tool on Waste Generation and Management”, Headline Project Report, Final Report for the European Commission DG Environment under Framework Contract No ENV.C.2/FRA/2011/0020’, http://ec.europa.eu/environment/waste/target_review.htm

52


Figure 12. Net social costs of each option compared to Option 1. Source: Impact Assessment of the proposal for reviewing the European waste management targets, Figure 42, p.87

Table 6 shows the detailed figures for financial and external costs, jobs and GHG emissions for the different options. It may be seen from the table that all the Options except 3.3 were shown to have negative financial and social costs (i.e. positive benefits), as well as leading to extra employment and reduced GHG emissions. The results of the case studies in EMInInn and other literature on the rebound effect raise the question as to whether the extra employment and reduced costs for waste management would lead to macro-economic expansion that would reduce the GHG reductions shown by the model. Not having a macro-economic component, the model in Figure xx could shed no light on this possibility.

53

Environmental Macro Indicators of Innovation Table 6. Costs and other outcomes from different options.

Source: Impact Assessment of the proposal for reviewing the European waste management targets, Table 12, p.88 Table 7 shows the scoring assigned to the options on the basis of these numbers across the various objectives of the proposal. Table 7. Scoring of the various options against the objectives of the proposal

54


Source: Impact Assessment of the proposal for reviewing the European waste management targets, Table 13, p.90 On the basis of these results the Communication recommended the adoption of Options 2 and 3.7.

4.1.3.2 Example 2: Impact Assessment of the Eco-Innovation Action Plan Background Eco-innovation is an integral part of the Europe 2020 strategy for smart, sustainable and inclusive growth. One important vehicle for it is the Eco-Innovation Action Plan (EcoAP), adopted in December 2011. The goals of the EcoAP are to boost innovation that reduces pressure on the environment and bridges the gap between innovation and the market. Seven key aspects of the Eco-Innovation Action Plan are as follows: 1) Using environmental policy and legislation to promote eco-innovation; 2) Supporting demonstration projects and partnering to bring promising, smart and ambitious operational technologies to market; 3) Developing new standards to boost eco-innovation; 4) Mobilizing financial instruments and support services for SMEs; 5) Promoting international co-operation; 6) Supporting the development of emerging skills and jobs and related training programmes to match labour market needs; 7) Promoting eco-innovation through European Innovation Partnerships Implementation of the plan is designed to be done in partnership between stakeholders, private and public sector, and the Commission. Impact Assessment The Impact Assessment (2011) 35 of the EcoAP had to be re-submitted to the Impact Assessment Board due to conceptual shortcomings relating to the problem definition, the analysis of drivers for eco-innovation, and the quantification of costs and benefits. This illustrates that the IAB has an important role in safeguarding and improving the quality of IAs, including the assessment of environmental impacts. The following section describes the main assessment methods used in the final IA. 36 The IA identified 5 policy options each containing a subset of actions. It is noticeable that the presentation of the assessment of each action is done in a very systematic manner with a high level of detail. Each action is assessed according to its contribution to achieving an objective and its impact on 1) the innovation system, 2) the environment, 3) competitiveness (see below presentation 35

http://ec.europa.eu/environment/ecoap/pdfs/comm_pdf_sec_2011_1599_f_en_impact_assesment_en.pdf Please note that the problem definition and the identification of objectives and policy options is not examined here, as the focus of EMInInn lays with the assessment of environmental impacts.

36

55


of assessment of option. Moreover the EU value added was examined and the estimated time until the actions take effect. The ambition to assess each action on such a high level of detail is without doubt the correct starting point, however, of even higher importance is the quality of the underlying analysis. Table 8 EcoAP impact assessment – option 3

In the following the methods used to arrive at the assessments are examined. The IA contains 5 dedicated annexes describing the assessment of each policy option. The following points are worth highlighting: •

•

It can be noticed that the assessments contain detailed qualitative consideration, but almost no quantitative estimates of environmental impacts (and other kinds of impacts). The Commission argues that this is justified by the fact that “many of the actions foreseen under EcoAP are indirect and, especially for the more strategic actions, impacts cannot be quantitatively assessed. Whenever possible, attempts to quantify effects are made, but given the complex relation between policy actions, final impacts, associated problems of attribution and additionally (as well as lack of evidence in some cases) it can be difficult to quantify effects”. Often were the quantification of impacts is not possible the Commission draws upon evidence from previous studies or experiences in other countries. For example for the assessment of Action 4 of Option 2 (“Develop and agree on performance targets for key 56


•

products, processes and services”) estimates of energy savings realised in Japan following a similar initiative are used as proxy. While the table above suggests that for every action a dedicated assessment of environmental impacts was conducted, such a practice cannot be identified based on a review of the relevant annexes in the IA. Often an generic assessment across the different types of impacts (innovation, environment, competitiveness) is provided (e.g. “impacts are expected to be…”, instead of “environmental impacts are expected to be…”). Therefore in certain cases the judgment appears rather random, as the generic and qualitative description of impacts does not really to allow for such a differentiated assessment as suggested by using the + and – signs. (e.g. option 1, action 1, p. 84)

Notwithstanding the points mentioned above, the crux of the matter is, as with all IAs, proportionality. It is not desirable to quantify environmental effects at all costs. The depth and scope of the analysis should be informed by the significance of expected impacts. EMInInn can play an important role in this respect as it provides new methods for measuring environmental impacts which potentially reduce the costs of quantitative assessments.

4.2 Monitoring Monitoring means observing (and measuring) the progress or quality of something over a period of time. In contrast to ex ante Impact Assessments and ex post evaluations monitoring does not aim to identify causal relationships between variables. Thus monitoring is conceptually much less challenging than the other two types of evaluations. In practice, monitoring and ex post evaluation are often intertwined as monitoring is basically a first step in ex post evaluation. 4.2.1 Requirements There are no separate binding requirements concerning monitoring in the EU. Instead the requirements are laid down in the Commission’s Impact Assessment Guidelines. The Impact Assessment Guidelines set out that part of every IA should also be to identify core progress indicators of the intervention and “a broad outline of possible monitoring and evaluation arrangements”. This is intended to help policy makers ”to check if implementation is ‘on track’ and the extent to which the policy is achieving its objectives”. The IA should also describe which data is needed and how it will be collected during monitoring. It should be noted that the monitoring framework policy makers have to develop as part of the IA will be subject to the review of the Impact Assessment Board just like all other parts of the IA. However, if the planned monitoring activities are actually carried out is not reviewed by the IAB. 4.2.2 Guidance Annex 13 to the Commission’s Impact Assessment Guidelines provides some further guidance on indicators, monitoring and evaluation. However, the information provided is rather generic and not specific to environmental monitoring. Nevertheless there is a wealth of data administered at EU level that can feed into environmental monitoring. The data is maintained among others by:

57


• • •

• •

•

The Joint Research Centre 37 - The Commission's in-house science service The European Environment Agency 38 - An agency of the European Union with the task to provide sound and independent information on the environment The European Environment Information and Observation Network 39 (EIONET) - A partnership network of the European Environment Agency (EEA) and its member and cooperating countries that supports the collection of data and the dissemination of information concerning Europe’s environment Eurostat 40 - The statistical office of the European Union The Eco-Innovation Observatory 41 - A project functioning as a platform for the structured collection and analysis of eco-innovation information, gathered from across the European Union The ODYSSEE MURE project 42 - A project of the 28 EU Member States and Norway which monitors energy efficiency trends and policy measures in Europe

4.2.3 Current practice In general it can be said that monitoring activities are less structured and controlled than Impact Assessments. Nevertheless, the importance of environmental monitoring on the macro-level has increased significantly in the last decade as the EU is setting more and more ambitious environmental targets. Consequently a rich pool of macro-environmental data has emerged; administered by organisation such as the EEA, the JRC, or the Eco-Innovation Observatory. On the level of individual regulations (e.g. Energy regulations) environmental monitoring appears to be less common and not systematic. The following section provides a detailed overview of current policy practices in the area of monitoring. The monitoring of environmental impacts, especially on the macro-level, was expanded during the last decade. This development is due to the increasing political priority given to environmental outcomes. The identification of quantitative environmental targets in the Europe 2020 Strategy is just one illustration of this development, which is rather extensive. 43 Consequently the importance of monitoring has risen as well, as monitoring is necessary to be able to track progress towards achieving environmental objectives. Examples of macro-level environmental indicators are: • • • • •

Reduction of greenhouse gas emissions (Europe 2020 strategy) Energy from renewables emissions (Europe 2020 strategy) Energy efficiency emissions (Europe 2020 strategy) Indicators within the Research Efficiency Scoreboard (part of the flagship initiative “A resource-efficient Europe”) Indicators within the Raw Materials Scoreboard (currently being developed)

37

https://ec.europa.eu/jrc/en/research-topic/environmental-monitoring www.eea.europa.eu 39 www.eionet.europa.eu 40 http://epp.eurostat.ec.europa.eu/portal/page/portal/statistics/search_database 41 www.eco-innovation.eu 42 www.odyssee-mure.eu 43 Environmental targets in Europe 2020: Reducing greenhouse gas emissions 20% (or even 30%, if the conditions are right) lower than 1990, 20% of energy from renewables, and 20% increase in energy efficiency 38

58


Another prominent over-arching monitoring framework is the Eco-Innovation Observatory. A principal objective of the Observatory is to inform the ‘Eco-innovation action plan’ (EcoAP). It provides detailed information on countries’ performance in eco-innovation in form of an EcoInnovation Scoreboard (database), country reports, and good practice examples. 44 Also the European Environment Agency (EEA) plays an important role in monitoring whether the EU is on track in achieving its environmental objectives. For example in 2013 it published an extensive study 45 on environment targets and objectives, including an assessment of the progress made in different policy areas. The study identified 128 targets across EU environmental and resource policies in nine areas (energy, greenhouse gas emissions and ozone- depleting substances, air pollution and air quality, GHG and air emissions in transports, waste, water, sustainable consumption and production and resource efficiency, chemicals, biodiversity and land use), of which 63 are legally binding targets and 68 non-binding strategic objectives to be achieved between 2010 and 2050 (most of them by 2020). Additional the EEA is a major source of information used by other parties for environmental monitoring. From a study of specific EU laws (Regulations, Directives) in the area of Energy and Climate Change (Milestone 19) it becomes apparent that monitoring is much less common and not systematic on the level of individual regulations. For example monitoring clauses are largely absent in the Energy Performance of Buildings Directive (EPBD) and Energy Efficiency Directive (EED). A study in the area of Resource and Waste policy (Milestone 18) illustrates that monitoring is also difficult because it often relies on the input of national authorities. Those might feel that data quality is not sufficient enough in order to establish monitoring systems even in front of quantitative targets, as it was the case with Germany concerning the implementation of the Waste Framework Directive (WFD). However, there are also many positive examples of countries establishing comprehensive monitoring frameworks concerning the WFD , such as the Netherland’s National Waste Management Plan. Another positive example of environmental monitoring in EU policy making is the European Strategic Energy Technology (SET) – Plan. The development of the SET Plan has been accompanied by the development of SETIS, the Strategic Energy Technologies Information System. This is a website that serves as a central information hub for SET-Plan activities, programmes and data. SETIS is managed by JRC’s institute for transport and energy, and has responsibility for a range of monitoring activities, including those relating to key performance indicators and the monitoring of financing. With regard to European Industrial Initiatives (EIIs), Key Performance Indicators (KPIs) represent an essential toolkit for monitoring and reviewing their overall progress, and of the individual research, development and demonstration (RDD) activities performed in frame of their implementation. In particular, KPIs have been defined with reference to the following EIIs: wind, solar, carbon capture & storage, nuclear and bio-energy.

4.3 Ex-post Evaluation Ex post evaluations have generally received less attention than Impact Assessments even though they are of similar importance in the policy cycle (and can be even more important if the effectiveness of typical problem-solving policies, as all those on the environment, is concerned). One 44 45

www.eco-innovation.eu http://www.eea.europa.eu/publications/towards-a-green-economy-in-europe

59


reason for this paradox is political: Newly elected politicians often want to initiate new projects to fulfill their electoral promises, and therefore dedicate less attention to the evaluation of existing policies. A more substantial reason is that the effects of a policy may require years to be fully deployed and many indirect (systemic) effects can arise thus making the analytical possibility of assessment lower than in the case of ex ante IAs. In addition, differently from certain economic and social policies, e.g. employment policies, that can have available extensive datasets of micro-data (individuals) to test for the effect of a specific policy, typically environmental policies have countries as units of reference with a short time series record, which prevents to apply conventional methodologies based on control-samples and counterfactuals. These methodological issues are vey similar to those faced by EMInInn in the specific area of environmental effects of innovation. The unlucky consequence is that, in many cases, ex post evaluation is meant as just the completion of the policy implementation process (e.g. Directive transposition and application at the different administrative levels) instead of a measure of the real-world effects of the policy on the problem at stake. This enduring poverty of instruments and performances is illustrated also by the fact that the current Evaluation Standards of the Commission are much shorter and less detailed than the Impact Assessment Guidelines. 4.3.1 Requirements The current Evaluation Standards date back to 2004 and set out standards in the following five categories: • • • • •

Resources and organisation of evaluation activities, Planning evaluation activities, Designing evaluations, Conducting evaluations, and Dissemination and utilisation of evaluation results

There is no reference to the assessment of environmental impacts (neither to the assessment of economic and social impacts). Instead the standards focus more on the procedural aspects of evaluation within the Commission. An important point to note is that in contrast to Impact Assessments the quality control of evaluations lays with the individual Directorates and not with a central oversight body as the Impact Assessment Board. The Commission is currently in the process of revising its Evaluation Standards. A public consultation on the draft Commission Evaluation Policy Guidelines ended in February 2014. The draft Evaluation Guidelines provide a first impression of how the future guidelines could look like. It can be expected that they will be more detailed and more focused on actual evaluation methods. However, just like the 2004 version they will probably not contain an explicit requirement to evaluate possible environmental impacts of an intervention, but focus on the following general evaluation criteria: • • • • •

Effectiveness Efficiency Coherence Relevance EU added value of the intervention 60


4.3.2 Guidance To support policy makers in implementing the Evaluation Standards a detailed “Practical Guide” 46 on evaluation within the Commission was published in 2004. The guide contains further explanation of the standards and provides a rich repertory of good practice examples. Annex E also provides an overview and a brief description of the most common evaluation tools and techniques (e.g. surveys, focus groups, descriptive statistics, econometric and input-output models). However, no detailed information on how these methods can be used to evaluate the environmental impacts of EU interventions is presented (the same is true for social and economic impacts). Only the “Guidebook to assessing environmental impacts of research and innovation policy” 47 (2013) published by Erawatch provides detailed guidance on ex post assessment. A number of sources that provide environmental data that might be used for ex post evaluation is outlined in section 4.2.2 4.3.3 Current practice Generally it can be said that while much environmental data on the macro-level is available, policy makers are not always aware of or not familiar with the different methods and models that can be used to measuring the ex-post impact of innovation (or technologies) on the environment. However, positive examples of evaluations examining the effect of innovation on the environment do exists, such as the evaluation of the CIP Eco-innovation market replication programme. Also the “Guidebook to assessing environmental impacts of research and innovation policy” 48, which was published by Erawatch in 2013, might help to improve ex post evaluation of environmental effects. The following section provides a detailed overview of current ex post evaluation practices. A review of a number of milestones produced through EMInInn (MS 18, 19, 20, 21) covering a range of EU policy areas (innovation policy, energy and climate policy, biodiversity policy, resource and waste) highlights several shortcomings in the current system. The most important results of the analysis are outlined in the following: •

•

•

A variety of methods are used for evaluation at EU level. For example for the evaluation of the ‘CIP Eco-innovation market replication programme’ participants had to fill in a survey involving questions about realised environmental gains. By and large, eco-innovation outcomes of policy are not directly assessed ex post, except to the extent that deployment of existing innovations constitutes an eco-innovation outcome. (MS 19, p. 22 ) An important success factor for evaluating environmental effects that was identified during the evaluation of FP5 and FP6 was that ex post evaluation must be supported by an monitoring and evaluation strategy established beforehand (meaning before a policy is implemented). Such a strategy facilitates the collection of relevant data and therewith provides the basis for a meaningful and reliable evaluation (Milestone 21, p. 27). Where

46

http://ec.europa.eu/smart-regulation/evaluation/docs/eval_activities_en.pdf http://europa.eu/!Xm84ND; summary: http://ec.europa.eu/research/evaluations/pdf/archive/other_reports_studies_and_documents/envti0413167enn_002.pdf 48 http://europa.eu/!Xm84ND; summary: http://ec.europa.eu/research/evaluations/pdf/archive/other_reports_studies_and_documents/envti0413167enn_002.pdf 47

61


•

Member States have to collect data themselves (e.g. in form of indicators) attention must be paid to ensure the comparability of information (see Milestone 18, p. 43). The European Environment Agency (EEA) appears to be on the forefront of monitoring and evaluating the environmental impact of EU policies (including innovation policies). An example of an analysis of extraordinary depth is the 2014 report on “Resource-efficient green economy and EU policies” 49. Chapter 5 provides an detailed assessment of the role of innovation in improving emission performance of selected countries and EU27 based on Structural Decomposition Analysis (see Box 2). The report also highlights the positive correlation between emission-intensity performance of countries (macro level) and the adoption of energy/emission innovations by companies in the country based on Community Innovation Surveys (CIS) data from 2006 to 2008 (micro level).

Box 2. Methodology used by the EEA to measure the role of technology and innovation in changing environmental pressures at the macro level Structural Decomposition Analysis (SDA) is used to decompose aggregate changes in time of CO2 emissions into various components by considering the input-output relationships between sectors (structure of the economy). The components considered are change in (i) emission intensity: (ii) technical change; (iii) structure of final demand, and (iv) level of final demand. Two different versions of SDA are developed for: (i) the 'production perspective', and (ii) the 'consumption perspective' that are usually adopted in an Environmentally extended input-output (EEIO) framework. In the production perspective (or production footprint), only direct domestic emissions are considered while final demand consists of total final demand (domestic demand and export) of domestically produced goods. In the consumption perspective (or consumption footprint), foreign emissions are also considered by including in the matrix of inter-industry transactions both domestically produced and imported intermediate inputs, while final demand includes overall demand by resident agents, thus including domestic and imported final consumption but excluding exports. The results show that emission intensity, the indicator closest to being a proxy for environmental innovation, plays a crucial role in controlling environmental pressures. Emission intensity has a larger role in controlling environmental pressures than other factors such as the level of demand and the changing industrial mix. After emission intensity, the second most important factor is technical change as represented in SDA by the coefficients of supply and demand linking the different sectors, which represents the technology of production of the economy.

4.3.3.1 Example: Ex-post evaluation of the CIP Eco-Innovation market replication programme Background The CIP Eco-innovation Initiative was launched in 2008 with the aim to support commercially oriented eco-innovation among European businesses. Funding is offered through annual calls for the commercialisation of market-ready new products, processes and systems within the private sector 49

http://www.eea.europa.eu/publications/resourceefficient-green-economy-and-eu

62


that have a positive environmental impact . The initiative is managed by the Executive Agency for Competitiveness and Innovation (EACI) in cooperation with DG Environment of the European Commission, and has a total EU funding budget of about €200 million for the period between 2008 and 2013. Applications for funding are invited under five thematic priorities, which are: materials recycling, sustainable building products, food and drink sector, water, and greening businesses. In the last call (2013) environmental benefits was one assessment criteria. 50 Applicants were also asked to provide a summary of quantified environmental impacts in a life-cycle approach. Ex-post evaluation In 2013 ICF GHK was commissioned to carry out an evaluation 51 of the results achieved by 126 projects under the CIP Eco-Innovation market replication programme between 2008 and 2010. In the following the main evaluation methods used in the study to assess environmental impacts are examined. The objective of the study was to answer eight specific evaluation questions, including the question “What have been the environmental impacts of the projects?” Information was collected through a desk review of project documentation, an online survey of project coordinators, and ten ‘deep dive’ case studies. Key performance indicators (KPIs) concerning environmental impacts were: energy savings, greenhouse gas emission reductions, waste savings, resource savings including raw materials and water savings. The assessment of environmental impacts of the project was done in five steps: • • • • •

Step 1 - Screening of indicator data and other environmental information supplied by projects to fine-tune methodology; Step 2 - General survey to obtain contextual information on the projects and establish main impact domains, thus enabling project categorisation; Step 3 - Customised survey of categorised projects to gather basic, standardised project flow data; Step 4 - Data processing, modelling and calculation of environmental impacts enabling a consolidated impact statement to be produced; and, Step 5 - Monetisation of impacts on a basis agreed with the Steering group using standard valuation approaches as applied to the consolidated statement of impacts.

During step 4 net environmental savings were calculated using Life Cycle Impact Assessment (LCIA). The savings were estimated in comparison to a baseline scenario, which project coordinators were asked to provide (e.g. which existing alternative products, processes or services would have been used without funding). It is noticeable that the study is very clear about the assumptions that had to be made to calculate environmental savings and how these assumption might influence results. During step 5 the impacts were monitised for each impact category based on conversion rates

50

http://ec.europa.eu/environment/eco-innovation/files/docs/getting-funds/2013/call-for-proposals2013_en.pdf 51 http://ec.europa.eu/environment/eco-innovation/files/docs/publi/report-eco-innovation-results.pdf

63


established in a Belgian study 52 (e.g. 1 square-meter of agricultural land used per year equals costs of 0,036 euro). The calculations allowed the study to provide estimates of 1) aggregate environmental savings for each KPI (e.g. tons of reduced CO2 emissions), and 2) total savings generated by all projects in monetary terms (see Table 9). The results are given for two different points in time: For the end of the project and for two years after the end of the project. The growth multiple, that is the number by which savings have multiplied between the end of the project an two years later, illustrates how successful the projects have been in diffusing/replicating environmental benefits in the market. Table 9. Estimated environmental savings from 125 projects of the CIP Eco-innovation programme

53

ICF GHK (2013, p. 91) based on analysis of Eco-innovation project survey responses and project data

Overall, the study is very innovative in actually trying to quantify environmental savings of a concrete policy intervention on EU level. The study is also exemplary in addressing methodological difficulties and uncertainties, and their effect on the results. The impact of mayor uncertainties (e.g. using data predominantly from Italian and Spanish projects) was even analysed through sensitivity analysis. A shortcoming of the study is that it examines a rather small sample. The calculation of environmental savings was based on data of only 48 projects, and grossed up to the total of 125 projects. The study also didn’t examine rebound effect on the macro level, due to a focus on LCA. In light of future evaluations of the programme the tools developed by EMInInn might be highly useful, as they provide possibilities to integrate bottom-up life-cycle approaches and top-down macro models.

52 53

OVAM (2012) - Milieugerelateerde Materiaalprestatie van Gebouwelementen

The monetised environmental gains across all eight KPIs is estimated as being approximately €40 million at the end of the project, rising to €833 million after two years (ICF GHK, 2013, p. 90) .

64


5 Identifying the environmental pressures from specific innovations Environmental policy in general, and eco-innovation policy in particular, frequently involves some technology-specific elements. As a result, considerable effort goes into technology assessment procedures that attempt to determine the expected environmental consequences of deploying or a particular technology or accelerating innovation in a technology field, and the costs of doing so. Technology assessment is used to inform prioritisation and funding decisions around research, development, demonstration and deployment spending. For example, technology-specific assessments are prominent in the analysis underpinning the European Strategic Energy Technology Plan (SET-Plan), and the UK’s Technology and Innovation Needs Assessment process. The review of current policy processes revealed a variety of practices with respect to technology assessment (see section 4, and Milestone 21). Policymakers rely on a considerable variety of tools and methods. In the field of low-carbon innovation, which has been a particular focus of ecoinnovation policy, energy-economy models have been prominent tools for informing energy innovation objectives and options. These procedures must negotiate a variety of challenges in assessing the environmental consequences associated with the diffusion of new technologies. This section 54 first provides an overview of the major challenges faced in assessing the full environmental consequences of particular innovations whether ex ante or ex post. The subsequent section highlights the tools developed and used within the EMInInn project, and discusses the strengths and weaknesses of each approach.

5.1 Challenges in technology assessment of supposed eco-innovations The challenge for attempts to understand the full environmental pressures arising from the development and diffusion associated with a specific technology lies in part in the diversity of ways in which production and consumption activities interact with both the environment and with each other. This section provides an overview of the key challenges for this sort of analysis, before providing an overview of the tools developed within EMInInn to overcome some of these challenges (in Section 5.2). 5.1.1 Pressures arising across the life-cycle of the product or service Products and services, even those that fulfil an identical function, may differ significantly in the stage of the life-cycle that generates the most significant life-cycle pressures. The point is easily illustrated with energy technologies: environmental pressures associated with solar photovoltaics (PV) are mostly associated with the manufacture and production stage, while those associated with a coalfired power station are largely attributable to the use stage. Failing to compare impacts across the full life-cycle can obscure impacts, and render comparisons between technologies unhelpful from a policy perspective (Figure 12). This is well known, but not always taken into account by technology assessment tools. For example, a major class of analytic tools widely used to inform energy R&D and energy policy decisions (i.e. energy system models) does not incorporate the full life-cycle of energy technologies. To be sure, the differences between power generation technologies are very largely determined by the use 54

Note that much of this section is reproduced from, or adapted from, D10.2

65


stage, and the other life-cycle stages are generally less important in determining overall pressures. However, in some cases these life-cycle emissions are significant (as with the upstream cultivation emissions associated with some bioenergy sources, and with PV installed in regions with low levels of sunlight (Bush et al., 2014))

Product-service system of interest

Production

Use

Disposal Figure 13. Life-cycle assessment.

5.1.2 Trade-offs between environmental pressures A technology that improves performance for one environmental pressure may generate worse performance for other pressures (Figure 13). For example, diesel vehicles are more efficient (reducing CO2 emissions per km compared to a petrol alternative), but also incur higher emissions of pollutants that are directly harmful to human health. Assessments that focus on a single pressure miss such dynamics. Current policy practice does not always take into account such dynamics in policy assessments. However, the review of policy practice noted an example in which the impact assessment process highlighted exactly this concern. The impact assessment process for the Renewable Heat Incentive (a UK policy) identified the potential for the policy to exacerbate air quality problems, through support for small-scale biomass heating systems. In response to the impact assessment analysis, the policy was changed.

66



Production

Use

Disposal Pressure 1 Pressures 2, 3, …

Figure 14. Life-cycle assessment with multiple pressures.

5.1.3 Pressures induced through economic feedbacks A major challenge in the assessment, either ex ante or ex post, of a supposed eco-innovation is the extent to which economic feedbacks may be taken into account. Increases in the consumption or production of a particular good or service will have multiple effects on other economic actors, supply chains and consumers, resulting in price and income changes. The idea of the ‘rebound effect’—in which energy efficiency improvements result in lower than expected energy savings because the effective price of energy services falls, stimulating demand for that service—is well known, and originated within energy economics. Within EMInInn, David Font Vivanco and Ester van der Voet have argued that the industrial ecology literature has greatly expanded this concept to incorporate a range of economic feedbacks associated with the diffusion of a good or service, and the environmental pressures that result (Vivanco and van der Voet, 2014). Failure to account for pressures induced through economic feedbacks can result in a misleading analysis of the effects of a technology or policy. The most well-known example is that of biofuels, and the concerns that policies to promote biofuels result in higher global food prices, with resulting induced conversion of land use to agriculture, more-than offsetting the emissions benefits of substituting petrol or diesel with biofuels. There are a wide variety of pathways by which these effects occur, and different tools account for different pathways. First, direct rebounds relate to the extent to which changes in the price of a good or service influence demand for that service (Figure 14). For example, an efficiency improvement in vehicle technology will reduce the cost-per-kilometre of driving, thus stimulating greater demand for driving.

67



Production Demand effect: Direct rebound

Use

Disposal Pressure 1 Pressures 2, 3, …

Figure 15. Life-cycle assessment with analysis of direct rebound effects

Second, as an innovation diffuses into market, there are substitution effects across the supply chain, as the innovation displaces direct alternatives, with knock-on effects throughout the sector. This is illustrated in Figure 15, in which the substitution effects are illustrated by showing the production cycle for a competing product or service. Some of these can result in a reduction in the environmental benefits of the diffusion of a technology. For example, an innovation that improves the efficiency of coal consumption in the power sector might lead to a reduction in power sector coal consumption, with a resulting fall in coal price – which might then lead to an increase in coal consumption in the industrial sector. Similarly, there may be changes in the demand for competing or related products, much as occurs with the direct rebound effect. These sector-wide equilibrium effects are captured by partial equilibrium models, including many energy system models. Note that Figure 15 shows a system boundary including partial equilibrium relationships along the full lifecycle, and including multiple environmental pressures, but this will not always be the case (e.g. MARKAL/TIMES models typically include only greenhouse gases as environmental pressures, and do not incorporate the full life-cycle, as discussed below).


Demand effect: Direct rebound

Sector-wide substitution effects

Production

Production

Use

Use

Disposal

Disposal

Pressure 1 Pressures 2, 3, …

Demand effect: indirect rebound


Figure 16. Sector-wide economic feedbacks.

68


Finally, the diffusion of an innovation can generate economy-wide substitution and demand effects, with changes in the demands for goods and services across the economy, and substitutions in their production (as illustrated in Figure 16). These whole economy interactions can be captured with a general equilibrium model.


Production Demand effect: Direct rebound

Sector-wide substitution effects

Production

Use

Use

Disposal

Disposal


Rest of the economy


Demand effect: macroeconomic rebound


Figure 17. Economy-wide economic feedbacks

5.1.4 Changing nature of processes that contribute to life-cycle pressures A challenge for analysis of the expected future (or past) environmental pressures associated with the diffusion of energy innovations is that technologies and processes change, both over time and as they are diffused. This can change the life-cycle inventory of the product-service system, either through substitution of inputs (which of course can occur in response to price changes induced by the diffusion of the innovation itself), or through changes to the processes. 5.1.5 Other system interactions: the importance of technological explicitness Technologies interacting in systems can influence the performance of other technologies within the system. For example, it has been argued that the presence of a significant share of intermittent renewable energy within an electricity power system could reduce the average efficiency of operation of fossil plant, by increasing the need for fossil plant to undergo large fluctuations in output. Similarly, the needs for peak margin capacity in a power system will differ depending on the mix of generating technologies. Analytic tools with highly aggregated representations of technology will overlook such interactions, while more detailed tools can explicitly assess their effects. 5.1.6 Spatial, temporal and functional boundaries In LCA, environmental pressures are calculated with reference to a functional unit, such as a vehicle kilometre. As a result, indicators of environmental pressure developed using this approach do not have a defined spatial boundary. However, where multi-regional input-output data is used in EEIO or hybrid IO analysis, it is possible to identify the regional incidence of environmental pressures 69


associated with a particular functional unit – though of course any future projections of this data would make bold assumptions about non-varying global production structures. The functional nature of the boundary for most LCA-based approaches is a weakness for forms of technology assessment that explicitly aim to address the extent to which a particular technology might meet national (or other geographic) targets. Temporal boundaries and assumptions can also be important. The pressures associated with the lifecycle of long-lived energy equipment, such as power stations or network infrastructure, will be generated over the course of decades, as will pressures associated with land-use change. Furthermore, the economic feedbacks associated with the diffusion of a given technology may take time (one can consider, for example, the economic changes facilitated by the steam engine). As a result, the answer to the question “is x an eco-innovation?” may depend critically on the time-scale over which the analysis is framed.

5.2 Tools for ex post and ex ante assessment of specific technologies developed within EMInInn The EMInInn project has developed and applied a number of analytic tools for examining the environmental pressures generated following the development and deployment of new technologies. This work has spanned technology fields of great relevance to European and Member State eco-innovation policies, including energy and climate change, transport, waste and buildings. This section provides an overview of those methods, highlighting their strengths and weaknesses, and points the reader to relevant deliverables that contain a fuller description of the analysis and the methods developed. 5.2.1 Ex post and ex ante scenario-based LCA and IOHA A simple approach to examining the potential life-cycle environmental pressures associated with the diffusion of a technology is to scale-up micro-level data developed on the basis of a life-cycle assessment. This is often done with scenarios, and this was the approach taken within WP8, for both ex post and ex ante assessment of the environmental pressures associated with the diffusion of energy from waste (EFW) technologies and waste recycling (at different geographical scales). Within the ex ante analysis, WP8 developed scenarios of alternative possible counterfactuals and diffusion paths for EFW technologies in Italy. This provides a sense of the potential for EFW to reduce environmental pressures, highlighting the implications of different policy options and trends (see D8.2). The analysis for Italy highlighted that a core uncertainty in the scale of environmental savings relates to the counterfactual electricity generating technology. When EFW offsets coal-fired power, environmental savings are considerably larger than when it offsets gas. Both WP8 and WP4 also applied a scenario-based approach to ex post technology assessment. WP4 applied a more sophisticated approach, using input-output based hybrid LCA, as described in Box 1. This approach, described in detail in D4.3, enabled an ex post examination of the contribution of

70


wind, PV, CCGT and flue-gas desulphurisation technologies to changes in environmental pressures between 1990 and 2010. In both cases, a key set of assumptions relates to the historical counterfactual scenarios. Just as with future scenarios, alternative possible pasts are subject to profound uncertainty (Bunzl, 2004): one cannot know what would have happened, for example, if flue-gas desulphurisation had not been invented. An attributional analysis (what environmental pressures can be specifically attributed to flue-gas desulphurisation) is relatively straightforward, and one can then compare the environmental pressures associated with coal power with flue-gas desulphurisation to coal power without it. However, this is not the same as an analysis that aims to identify what would have happened, which requires a consequential assessment. While economic relationships can be used to estimate more probable outcomes, the observed lack of predictive power of energy-economy models (Smil, 2000) suggests that any such estimates should be treated with caution. The implication is that it is challenging to make confident assessments of the full consequences of diffusion of a particular technology. Using a variety of plausible historical counterfactual scenarios, however, provides some sense of the implications of the uncertainty, just as in ex ante analysis.

Box 1. Indirect emissions in Process LCA, Environmentally-extended Input-Output Analysis and Hybrid Analysis 55 Two basic approaches have been traditionally used to assess the indirect environmental pressures associated with a product (group) or service: process-based LCA and EEIO analysis Process LCA is a bottom-up approach that estimates the environmental pressures or impacts that arise from across the whole life-cycle of a product or service (from resource extraction, through production, use and finally end-of-life disposal). It is based on engineering analysis of the processes and materials involved, and represents the processes in great detail. Nonetheless, it fails to cover the whole value chain in its system boundaries, which leads to truncation errors. The resulting life cycle inventory (LCI) gives, among others, information on the emissions cumulated across all the lifecycle stages of a reference product from cradle to grave without any time boundaries. Environmentally-extended input output analysis is a top-down tool that is commonly used to estimate the direct and upstream environmental pressures from a final consumption perspective in a given year. In most cases it is based on a combination of monetary IO tables, which depict the interdependencies between the production (sectors), investment and consumption (e.g. households, government, etc.) units in an economy, and environmental extensions, which give information on the direct environmental pressures exerted by economic sectors. These two elements are the core of EEIO tables. Thus, the environmental pressures cumulated along the production chain of products and services for final use are estimated by re-allocating the direct pressures to the end consumer based on the economic flows that trace the whole value chain of a product or service. In contrast to LCI-based emissions, the emissions from EEIO analysis refer to a single year and only allocate the pre-consumer stages to the reference product group or service. More differences in terms of the

55

This box is reproduced from EMInInn deliverable 4.3.

71


treatment of capital, post-consumer stages, transport activities, etc. in LCA and EEIO analysis are given in Usubiaga et al. (2014). Hybrid approaches have emerged more recently to overcome some of the limitations of process LCA and EEIO analysis. Depending on how the foreground and background systems are connected to each other, Suh and colleagues (Suh et al., 2004) define three types: tiered hybrid analysis, inputoutput based hybrid analysis and integrated hybrid analysis. In the context of the EMInInn project we have used input-output based hybrid analysis (IOHA), which is described in more detail by Suh and Huppes (Suh and Huppes, 2005) and by Usubiaga and colleagues (2014).

WP7 took a similar approach to the ex post assessment of innovations in the built environment, but used a simulation model to estimate the use-phase environmental pressures associated with energy efficiency technologies. This approach combined LCI data from existing databases with modelled environmental savings.

5.2.2 Dynamic IPAT-LCA with environmental rebound effect (DILER) 56 Through the DILER approach, product level estimates of innovations are modeled through time and scaled-up to the macro-economic level using innovation diffusion data. By calculating counterfactual scenarios in which the introduction of a given innovation did not take place, one can assess whether the innovation led to a decrease in environmental pressures at the macro level, an thus whether it can be regarded as an eco-innovation or not. For the temporal modeling exercise, technology change information will be gathered for each innovation when available. According to the IPAT-LCA 57 approach, eco-innovation indicator results can be described in terms of the contribution of a selected innovation to the change in a given environmental indicator, using decomposition techniques 58. The resolution of the change can be yearly or from a certain time period, depending on the innovation diffusion/technology change data availability. Following is presented an overview of the proposed method (Figure 17).

56

This discussion is reproduced from EMInInn deliverable 6.2 IPAT refers to the IPAT identity, in which Impacts = Population x Affluence x Technology. LCA refers to Lifecycle assessment. 58 Particularly Logarithmic Mean Divisa Index (LMDI) decomposition, as discussed further in EMInInn deliverables 6.1 and 6.2. 57

72


Figure 18. Overview of the IPAT-LCA approach

To increase the “behavioral realism” of the eco-innovation indicators, the environmental rebound effect (ERE) will be calculated and attributed to the corresponding innovations. In the context of this report, the ERE is defined as the environmental outcomes of a change in individual total demand as a result of a change in income due to price differences in providing a comparable function. For instance, if a transport innovation (e.g. a car with a new propulsion technology) providing the same function (e.g. passenger transport) as the relevant alternative (e.g. car with comparable propulsion technology) entails a reduction in travel costs, the ERE equals to the environmental pressures from the extra demand created due to the liberated income. The extra demand can be thus attributed to the technological characteristics of the innovation and so their associated environmental pressures. By establishing such a causal relationship, one disregards the contribution of systemic aspects, such as the fuel prices or the market structure. The ERE primarily differs from the traditional energy rebound effect in that various environmental aspects are studied instead of energy use alone (Font Vivanco et al., 2014b, 2014c). This differentiation has implications not only on its representation (multiple environmental indicators), but also on its definition, For instance, one can study the rebound effect from technical changes that are not strictly aimed at increasing energy efficiency. This aspect allows for broader analysis in the context of transport. For example, it allows to study transport innovations that are mainly aimed at reducing GHG emissions or other environmental issues. The ERE can be calculated through the model described in deliverable 6.3. This model uses LCA scores from specific innovations as a starting point and adds to these the calculated environmental pressures estimates related to changes in the total functional output (consumption basket) as a result of changes in effective prices from the use of the assessed innovations. The ERE model combines econometric analysis to model changes in demand, following similar approaches such as those from Chitnis et al. (2012) or Thomas and Azevedo (2013) to calculate the direct and indirect effects. A general overview of this model is described in Figure 18. Once the LCA results incorporate the ERE, they are scaled up through the dynamic IPAT-LCA model. We label the combination of the dynamic IPAT-LCA model with the ERE model as the DILER model.

73


Figure 19. Overview of the environmental rebound effect (ERE) model. E3IOT refers to environmentally extended inputoutput tables; HES refers to Household Expenditure Structures; ISTCM refers to ‘income-shifting with technology choice model’. Further explanation of the method is found in D6.3, and in (Font Vivanco et al., 2014).

The calculation of the ERE also allows for the appraisal of the multidimensional contribution of a particular innovation to an environmental indicator in terms of both technology and demand. For this, we follow the approach presented in task 3.3, which is based on index decomposition analysis. The proposed DILER model thus offers the following advantages: • • • • •

It keeps the high technology detail and life cycle approach of LCA. It offers behavioral realism by incorporating overall changes in demand due to changes in prices. It describes environmental pressures from single innovations at the macro-economic level. It incorporates technological change and offers temporally dynamic results. By linking technology and demand via the ERE, it permits to calculate the aggregate contribution of an specific innovation to changes in environmental pressures in terms of technology and demand.

An overview of the proposed methodology is presented following:

74


Figure 20. Overview of the proposed methodology. Highlighted in green, the information fully or partially provided in deliverable 6.1.

5.2.3 Soft linking input-output based hybrid LCA with energy system models Bottom-up energy system models provide a detailed depiction of the energy system, with explicit representation of the processes of primary extraction of energy resources, processing and conversion, delivery to consumers, and end-use. Well-known examples include the MARKAL/TIMES framework, widely used by governments to inform energy policy 59, including R&D and innovation policy 60. Such models account for some emissions associated with upstream extraction (fugitive emissions, for example), and they account for the efficiency losses and energy inputs associated with conversion and processing (e.g. in refineries and power stations), with transmission and distribution (losses in electricity transmission lines, energy use in fuel distribution, etc.), and efficiency losses in end use devices. Such models are “demand-driven” in the sense that the energy service demands across the economy are a key exogenous input into the models. Energy service demands associated with residential consumption, transport, industry and service sectors are all key inputs. However, energy system models do not directly account for the emissions associated with the manufacture of energy technologies. Moreover, various upstream processes are frequently not incorporated into the model in an explicit way (such as emissions associated with the cultivation and transportation of primary bioenergy resources), nor are indirect emissions associated with the operation and use of energy technologies. Instead, the activities that produce these emissions are located elsewhere in the model (e.g. in the industry sector for activities producing energy 59

Note that PRIMES has many of the same features of bottom-up energy system models discussed here, but it includes a greater representation of economy-wide feedbacks. Like MARKAL/TIMES, however, it does not enable optimisation of infrastructure-related emissions. 60 See, for example, the Strategic Framework of the UK’s Low Carbon Innovation Co-ordination Group, which was informed by both MARKAL and ESME, a similar model.

75


technologies; or in the agriculture sector for the cultivation of bioenergy), the energy service demands of which are specified exogenously. Yet many low-carbon energy technologies are more capital-intensive than their high-carbon alternatives, and structural shifts to low-carbon technologies (such as wind, for example) might be expected to lead to increases in activity and hence emissions from the industrial sector relative to the case in which fossil fuels continue to be used. Such endogenous changes in industrial production implied by energy transition scenarios are ignored by energy system models. Linking upstream emissions factors derived from IOHA to energy system models is one way of testing the importance of this assumption, and this is the approach that was developed within EMInInn WP4. The approach developed within EMInInn is fully described in D4.3, and the method is not further elaborated here. The important point, from a policy perspective, is that the model’s assessment of the relative importance of key technologies does indeed change when indirect emissions are taken into account. 5.2.4 Environmentally extended CGE Like EEIO-based approaches, some environmentally-extended CGE models provide good representation of the upstream emissions associated with relatively aggregated product groups and services for final use, across a range of environmental pressure categories, through the underlying environmentally-extended IO table on which they are based. They also incorporate strong representation of economic rebound effects, and can be considered more analytically consistent than a partial equilibrium or scenario-based approach (Rajagopal, 2014). These approaches incorporate direct and indirect emissions alongside the full production cycle. In recent years, EEIO tables have been used to develop environmentally-extended CGE models, including the EXIOMOD and GTAP-E frameworks used within EMInInn. These bring together a systematic framework for representation of the macro-economy with a full emissions accounting across the whole production value chain through the environmental extensions—representing emissions across the economy rather than simply arising from fuel combustion in the power sector. Ongoing work within the EMInInn project is testing the EXIOMOD approach to net GHG emissions associated with wind, PV and other power sector technologies. This approach showed how the wider potential effects of specific technologies could be assessed within a CGE framework, incorporating a wide range of economic feedbacks, and reporting results across a range of environmental pressures and economic outcomes. 5.2.5 Ex ante assessment of rebound potential At the micro-level, indicators can be used in conjunction with technology assessments to describe the potential of a given innovation to generate rebounds. The potential for a given technology to generate rebound effects appears to be rarely assessed within technology-specific innovation policy (such as the SET-Plan). However, where some data exists on demand elasticities, it is possible to estimate the potential for rebound, since the relative price of an innovation and its direct alternative are typically information that is already collected in the course of technology-specific policymaking. This information can be used to assess the resulting change in the price of an energy service, which can be used in combination with elasticity information to identify the direct rebound potential.

76


Indirect and economy-wide rebounds can be expected to be more difficult to estimate. However, D6.2 provides one way of screening innovations with respect to existing data, which may provide a way of identifying innovations with greater or lesser expected indirect environmental rebound effects.

77


6 Monitoring eco-innovation at the macro-level: applications and methodological advances in EMInInn Policymakers monitor macro-environmental trends as part of basic activities to assess the ongoing state of the natural environment, as described in Section 4.2. One of the aims of EMInInn was to contribute to policymaker capacity in identifying the contribution of innovation to changes in environmental indicators. In this regard, EMInInn has yielded: (i) insights from existing tools; (ii) data to support the use of existing tools, and (iii) the development of new tools for monitoring ecoinnovation at the macro-level. 61 The principal purpose of these tools is to examine macro-level trends in environmental pressures, and to isolate the effects of technological change, with no immediate insight into the specific technologies or changes occurring. However, note that several of these tools can also be used to identify the contributions of specific innovations to macro-level environmental outcomes, where micro-data is available. Indeed, some of the technology-specific tools discussed above (particularly the DILER approach) make use of the analytic approaches discussed here. 6.1.1 Index decomposition analysis Index decomposition analysis (IDA) provides a way of identifying the contribution of various factors to a change in an overall indicator. For example, one might seek to understand the relative contributions of technological change and sectoral structure on economy-wide energy efficiency. Decomposition analysis is suited to this role, and is already widely used in identifying the innovation contributions to changes in environmental pressures at the EU level. EMInIn did not attempt to advance the state-of-the-art with respect to this tool, but made use of it with a novel dataset to identify the contribution of FGD to sulphur emission reductions. Within EMInInn WP4, the EEA decomposition of sulphur emissions (EEA, 2006) was further developed by developing a new database of coal-fired power stations, in an attempt to identify the relative shares that could be attributable to a specific technology (Flue Gas Desulphurisation) and to a range of other, unidentified innovations and other changes. Officials at the EEA have requested a copy of the EMInInn coal and SO2 database, which is now being transferred to EEA for use in their ongoing SO2 monitoring activities. 6.1.2

Structural decomposition analysis

Structural decomposition analysis is defined as “the analysis of economic change by means of a set of comparative static changes in key parameters of an input-output table” (Rose and Chen, 1991). For (eco-)innovation analysis, SDA can decompose the change over time of an environmental stressor into its main independent contributing factors or determinants based on data from environmentally extended input-output (EEIO) tables (Hoekstra, 2005; Rose and Casler, 1996). Following the general formulation of EEIO analysis: 𝐺𝐺 = 𝑔𝑔 (𝐼𝐼 − 𝐴𝐴)−1 𝑦𝑦 61

Note that much of this section is adapted from, or reproduced from, D10.2

78


(where G represents total environmental pressures, g the direct pressure intensity, (I-A)-1 is the Leontief inverse (L) and y the final demand of a country), technological change or innovation can be decomposed into various contributing factors. If we were to decompose a change in G into the elements shown in the formula above, innovation could be represented as the sum of the intensity and the production structure effects. The combination of these two effects with the contribution of the final demand y explains the changes in the overall footprint of a country. In this case, the intensity effect indicates how variations in the environmental intensity (represented as direct pressures divided by unit of output) of the country or sectors have contributed to the overall change over the period under assessment. The second effect shows the aggregated contribution of changes in the production structure of the economy (represented as economic multipliers). Nevertheless, the contribution of (eco-)innovation can be assessed in more detail. For instance, the production structure effect can be further decomposed into sectoral technological developments in the form of changes in the direct input coefficients matrix and variations in product mix within each sector (Miller and Blair, 2009). Some authors have gone a step further and disaggregated the effect of the input coefficient matrix to assess the change in a sector’s intermediate input intensity, the average substitution of intermediate inputs between sectors, and the remaining variations not explained by the previous two effects by using the RAS technique (van der Linden and Dietzenbacher, 1995, 2000; Dietzenbacher and Hoekstra, 2002). In the ex-post assessments carried out in EMInInn, SDA has been used both in WPs 4 and 7. In WP 4, changes in the direct and indirect material use and air emissions induced by the domestic final consumption of electricity have been decomposed into their main determinants. In that case, an aggregated innovation factor that combines the intensity and production structure effects has been used. The domestic final demand of electricity has also been disaggregated into its structure, thereby showing changes in the mix, and the domestic electricity demand. In WP7, the intensity, production structure and final demand effects have been assessed separately. 6.1.3 Heat mapping changes in eco-efficiency Environmentally-extended input-output tables, if available over a time series, provide a straightforward means of identifying sectors where eco-innovation (measured as an improvement in eco-efficiency) is taking place, and where it is not. Following the methodology outlined in D3.2, it is possible to develop heat maps, showing changes in eco-efficiency of an industry with respect to specific environmental pressures over time (or, changes in eco-efficiency of a set of industries with respect to a given environmental pressure). This procedure, illustrated in Figure 20 provides a way of visually screening for the presence or absence of eco-innovation, either from the perspective of a particular sector, or from the perspective of a specific environmental pressure.

79


Figure 21. Heat map showing eco-efficiency of the electricity, gas and hot water supply sector with respect to eight environmental pressure indicators

This screening and monitoring tool can provide an early assessment of the macro-level ecoinnovation changes within a particular sector. Further, sector-specific analysis would clearly be required to identify the contributions of any specific innovations. Note that this approach does not reveal the extent to which innovation may be contributing to changes in demand or relative sectoral output. 6.1.4 Econometric and statistical analysis Statistical and econometric methods can be used to identify factors underpinning change in environmental pressures. In WP2, statistical analysis was used to identify whether changes in macroenvironmental indicators could be explained by the diffusion of particular innovations, with a focus on energy industries (wind and CCGT), road transport (diesel cars, and catalytic converters) and waste. This highlights the extent to which direct eco-innovation measures (in this case diffusion of ‘clean’ technologies) correspond with indirect measures (such as changes in eco-efficiency). The key insight (from D2.3) is that eco-efficiency changes of sectors depend on a more than the diffusion of a single innovation. Eco-efficiency changes are indicative of innovation, but a deeper analysis is required to determine the contribution from specific innovations and other factors. Econometric analysis – which can also be seen as a form of decomposition in a probabilistic statistical setting - has been used in WP5 on ICT and internet to study the relationship between ecoinnovation adoption and ICT innovation adoptions at the company level. Econometrics has been extensively used in WP8 to highlight the role of waste policies vis à vis other non-policy drivers in explaining the macro-scale shift of waste management away from landfill towards recycling and energy recovery observed in Europe. This is equivalent to test for the specific role of policies in pushing the waste system to move up along the EU waste hierarchy thus getting the expected net benefits of environmental pressures associated to higest levels of the waste hierarchy itself. To this 80


end, specific poliy indicators differentiated by country have been elaborated and used. Econometric techniques have also been used to test for if and how waste policies can influence the dynamics of the waste system in a direct way or indirectly through innovation. The same techniques have been used to test for the joint role of government-supported scrapping schemes and product innovation in the automotive sector for the accelerated scrapping of end-of-life vehicles. Finally, statistical techniques (cluster analysis) have been used to depict the evolving clusters of different local (province level) approaches to municipal solid waste management in Italy. Given its features, econometrics can provide measures of correlation and even causation that can be used in the ex post evaluation phase of the policy cycle.

6.2 Indicators for monitoring eco-innovation Indicators measure a certain state of affairs, and enable policymakers and other actors to monitor trends. By summarising and simplifying complex information, indicators ‘make visible’ complex phenomena. They can be differentiated from simple data or statistics by the presence of some underlying conceptual framework (Lehtonen, 2015). Indicators are descriptive, and do not establish causal effects, though of course indicator sets are frequently used as the basis for analysis that aims at establishing causal relationships. Policymakers seek indicators of eco-innovation for a number of reasons: • • •

•

to monitor eco-innovation progress, and so potentially identify sectors or environmental pressures that require further attention to understand and evaluate the success of specific programmes supporting eco-innovation to assess which nations are leading in terms of seizing green comparative advantage (by successful commercialisation of green technologies), and which nations are leading in terms of successfully deploying greener technologies and decoupling growth from environmental degradation to analyse the drivers of eco-innovation and its economic and environmental consequences.

81


Current indicators in use

Incentives and barriers: market opportunities, technical possibilities, institutions, etc.

Indicators of policy; knowledge stock and market opportunity; R&D budgets

Innovation processes Invention Diffusion

Policy Responses

Patents; product launches, market shares, etc

Economic feedbacks Price per unit service

Environmental pressures

Absolute and relative pressures

Figure 22. Currently used indicators describing the relationship between innovation, policy and environmental pressures.

The complexity of the relationship between innovation processes and environmental outcomes was illustrated in simple model in Figure 9. This same simple model is shown again in Figure 21, this time showing where indicators are currently used to monitor the complex relationship between innovation, policy and environmental pressures. The complexity of the relationship shown in the figure suggests that a single over-arching indicator of eco-innovation would not meet the needs of policymakers. Rather, indicators are required that help policymakers to monitor and understand the way in which the relationship between policy, innovation activities and environmental outcomes are co-evolving. No single indicator can satisfy all of these information needs. Most obviously, country A can successfully develop and commercialise an eco-innovation that might then be most widely deployed in country B, which then experiences the greater reduction in environmental pressures. Which is the more successful case of eco-innovation depends on what aspect of the innovation process policymakers are more interested in. Moreover, eco-innovation is a complex and systemic process, and many policy questions will not be best informed by a single all-encompassing metric. Indeed, a single composite metric might obscure more than it revealed, by appearing to reduce ‘ecoinnovation’ to a simple scalar quantity. This suggests that a range of eco-innovation indicators is required, along the various different “routes of impact”. This would require indicators that cover stimuli and input measures (relevant policies, market opportunities), output measures (innovation processes including invention and diffusion), rebound and other feedback effects, macro-environmental effects (direct and indirect environmental pressures). As demonstrated in Section 8.2, the stringency of environmental policies 82


(and particularly pricing policies) are important determinants of the full environmental consequences of eco-innovation, and are therefore relevant for indicator systems that cover the full system that drives eco-innovation at the macro-level. Many existing indicators might contribute to an ideal system of indicators for the eco-innovation process. Monitoring of changes in eco-efficiency across sectors (as illustrated visually in section 6.1.3) can act as a monitoring tool for identifying cold and hot spots. Subsequent analysis, much as currently performed by the EEA in the analytic reports that accompany online publication of environmental indicators 62, can identify the distinct contributions of specific factors, including innovation. Recent advances in tagging patents with ‘green’ patent codes has greatly facilitated the use of patent data for eco-innovation research. However, there are a number of areas in which indicators are currently rather sparse. In particular, there is a lack of indicators of green R&D spending, particularly in the private sector. This was highlighted by the MEI project (Kemp and Pearson, 2008). It is unfortunate that recent moves to harmonise European data gathering on the environmental goods and services sector have not included attempts to harmonise ‘green R&D’ data. The absence of data reduces the ability of policymakers to: A) determine the extent to which R&D focused on greener technologies does actually deliver either “supposed” eco-innovations (measured by green patents) or contribute to “revealed” eco-innovation (revealed by changes in sectoral environmental performance, though noting that this wouldn’t take economic rebound effects into account); and B) determine whether public R&D and private R&D priorities are diverging, as has been suggested in the context of energy research (Rhodes et al., 2014).

62

Currently published online here: http://www.eea.europa.eu/data-and-maps/indicators

83


7 When does the innovation process drive reductions in environmental pressures? Drivers, barriers and framework conditions of eco-innovation This section summarises the information arising from EMInInn on the factors and conditions under which innovation diffusion processes generate environmental benefits. Section 3.1.4 provided a summary of the main messages that can be identified in the eco-innovation literature regarding drivers, barriers and framework conditions. In section 3 barriers to (eco)innovation were defined as a factor exerting a negative impact on the (eco)innovation process (Piatier, 1984) whereas factors with a positive influence were called drivers, (Hadjimanolis, 2003). Framework conditions refer to those elements surrounding the innovation, contributing to its upscaling (e.g. for prototypes or pilots) or diffusion (e.g. for early market innovations). To our knowledge, the analytical approach of EMinInn in addressing these factors represent an original contribution to the broader literature. This is because EMInInn explicitly links eco-innovation diffusion and sustainability literature (Markard et al., 2012), advances the state of the art in terms of profiling and measurement (Kemp, 2010; Ekins, 2010) and attempts to provide a quantitative account of a macro-environmental and rebound effects (van den Bergh, 2013, 2011). This section discusses the drivers and barriers as identified in the analytical case studies of EMinInn, followed by evidence on the role of framework conditions – and it does so while noting that much of the existing evidence takes for granted that ‘green technologies’, ‘green R&D’ or ‘green patents’ are environmentally beneficial. It also highlights that the environmental consequences of a given innovation may depend rather strongly on policy conditions, such as the stringency of environmental regulations or pricing.

7.1 Drivers and barriers of (eco) innovation The Description of Work of EMInInnn originally envisaged a research activity with a sole focus on identifying particular technologies and creating fiches mentioning factors having allowed its diffusion across the European Union (EU). However, the research team considered that this action would not necessarily provide sufficient analytical insights about what factors contribute to a positive macro environmental impact of innovations. For this reason, the work of EMinInn was carefully modified to identify under what conditions a supposed eco-innovations would constitute a real eco-innovation. Hence, a central aspect of the analytical case studies (energy, ICT, transport, built environment and construction and waste management) was the identification of specific drivers and barriers in the different sectors and the use of a counterfactual scenario to determine its (positive or negative) environmental impact. As noted in section 3, drivers and barriers to eco-innovation can take a wide variety of forms. As in any other scientific research endeavour, the departure point for each case study was the revision of available evidence on external and internal factors driving or hampering eco-innovation (diffusion). A few general observations from the literature review are hitherto provided (see section 3.1.4):

84


•

•

•

A general message from the wider innovation literature (c.f. Taylor et al., 2005) suggest that expectations concerning future regulation and future prices can be important in driving innovator behaviour. Notwithstanding, most of the available literature reviewing or analysing the drivers (and barriers) of eco-innovation have a strong focus on manufacturing sectors and/or refer to consumer products (e.g. see factors cited in section 3.1.4). A large number of empirical studies provide no explicit differentiation between drivers and barriers for eco-innovation development, diffusion and use; moreover, just a handful of empirical (econometrical) studies analyse internal factors for eco-innovation. Most notably, there is an insufficient analysis of resources, capacities and competences for eco-innovation (c.f. Diaz Lopez, 2008, Del Rio et al., In press).

The empirical review of drivers and barriers to eco-innovation in the energy, ICT, transport, built environment and waste sectors constituted an important component for the overall analysis of the environmental effects of the diffusion of past innovations. The underlying assumption was to enquiry what would have happened to the environment if a particular innovation would not have been diffused. In doing it so, a few considerations and limitations of EMInInn’s empirical approach are worth to be noticed: • A variety of qualitative case studies across several other work packages (D4.1, D5.1, D6.1, D7.1) highlighted important roles for regulation, market demand, consumer preferences, etc., and the synergies between them in creating both drivers for and barriers against ecoinnovation. • EMInInn did consider the different stages of the innovation (diffusion) process and addressed (economic) sectors outside the manufacturing and service sectors (see Figure 9). • The scope of study of EMInInn focused on (economic) sectors and the diffusion of specific technologies; and not on firms developing or adopting them. Hence, the analytical case studies focused on external drivers and barriers for the diffusion of eco-innovation. 63 Albeit some of the evidence provided in this report use data from patents and other (innovation) input sources 64, EMInInn did not analyse drivers for and barriers for eco-innovation diffusion from the perspective of those actors developing or adopting the technology (firms or R&D centres). Hence the evidence provided in this report do not deal with firm’s (or managers) motivations, capacities or behaviour. EMInInn had a focus on the diffusion of particular technologies. As a result, the analytical /work implicitly focused on functions derived from the diffusion of innovation, for example providing a service (e.g. provision of internet, incineration), producing a commodity (e.g. power companies), manufacturing a good (e.g. electric car makers), or purchasing a good (e.g. housing), and the macro environmental effect (and rebound effects) that the diffusion of particular innovations would entail. Broadly speaking, the overall methodological approach for the analysis of cases followed the logic enounced below:

63

The only exception was the built environment and construction that took into account capabilities for innovation in the sector. 64 Refer to section 3 for an overview of common data sources and indicators for the measurement of eco-innovation. See also Kemp (2010).

85


1. Definition of the boundaries of the system, and characterisation of the main sector and technological features related to innovation and sustainability 2. Definition of a useful classification of (supposed) eco-innovations 3. Identification of factors hampering or contributing to the diffusion of selected innovations 4. Interaction, synergies and/or opposing effects among identified drivers and barriers. 5. Provision of main lessons and implications from analytical case studies. The points above are used to structure the remainder of this section illustrated with relevant information derived from the study of the diffusion of innovations in the energy, ICT, transport, built environment and waste sectors. 7.1.1 Boundaries of the system; sector and technological characteristics The first analytical step was defining the boundaries of the system and to clarify the research question. In most cases, the analytical boundaries were set at the sector and technology level. As noted in section 3, each analytical work package used a systems-based approach to innovation as analytical framework. The following theoretical underpinnings can be mentioned as examples of the use of a systems approach to innovation informing the analytical work of EMinInn: • Renewable energy: sector and technological innovation systems, mostly focusing on the role of regulation and markets. • ICT: technological innovation systems and the multi-level perspective providing an overview drivers and barriers at the landscape and meso level (internet innovation system) • Transport: an innovation systems approach guides the characterisation of system dynamics, in particular when considering trade offs between technology choices and environmental performance. • Built environment and construction: drivers and barriers arranged according to a market and system failures framework • Waste managed used a sector and regional approach to eco-innovation, with a strong focus on the role of regulations, markets and proximity. Understanding sector and technological characteristics of the innovation system and multi-level interactions / dynamics of the innovation process was the departure point of each case study. Providing a basic innovation and sustainability profile of sectors would enable a more detailed characterisation of innovations, for example: 1. The presence of strong lock-in and path dependency effects, particularly important role of economies of scale, capital intensiveness, significant network externalities for many systems and low R&D intensity are core characteristics of innovations in the energy system. 2. Internet is hardly ever associated to environmental gains, but mostly on energy efficiency. Diffusion effects on the environment can be observed when taking into consideration the different layers of the internet system and the multi-level factors driving innovation and change. 3. Transport: innovation (compared to relevant alternatives) is highly determined by external factors enabling diffusion and (supposed) environmental benefits, in particular the role of standards, regulation, public procurement and consumer preferences 4. In the case of the built environment and construction, compared to other sectors, it was noted that innovation within construction firms is minimal and innovation across the sector is 86


fragmented (e.g. incorporating innovations from other firms is practically non-existent). The sustainability imperatives in this sector are notable; therefore, the expectations for a wave of future green innovations in the sector. Hence, innovation is being driven by the imperative to achieve resource and energy efficiency. 5. Waste management considers the macro-level eco-innovation observes a very specific pattern of adoption/diffusion of innovations that reduce environmental pressures (compared to landfill), which can be observed in the EU during the last 20 years of implementing regulations in this sector. 7.1.2

Classification of eco-innovations

The second analytical was defining the unit of analysis that would be manageable and would explain with sufficient detail the different characteristics and features of selected innovations. This step was implemented because available classifications of eco-innovations would not necessarily provide a satisfactory level of disaggregation that could sufficiently allow a detail account of sector-level and technology-level characteristics of eco-innovation. Table 10 Selected innovations in EMInInn distributed according to the MEI classification of eco-innovation Eco-innovation general typology (MEI classification) A - Environmental technologies B - Organisational innovation for the environment C - Green products and services

D - Green system innovations E – Generic purpose technologies*

Selected innovations – analytical work packages Pollution control technologies (flue-gas desulphurisation, catalytic converters), waste management equipment (material recycling and composting, incineration and energy recovery, landfill), green energy technologies (wind, solar PV) NA. Services that are less pollution and resource intensive (car sharing, public bike schemes, park and ride facilities), green energy services (electricity from wind energy, electricity from solar PV), energy efficient products (energy saving light bulbs / LED’s / Fluorescent lights, high efficiency boilers), greener cars (battery electric vehicles, fuel cell vehicles) High speed rail systems, district heating Transport technologies (diesel engines, direct fuel injection systems), internet innovations(mobile phone, smart phone, wireless networks, data centres, electronic product re-use, long distance tourism), construction technologies (insulation of floor, wall and roof / loft, multiple-glazed windows), energy technologies (ground source heat pumps, combined cycle gas turbines) Co-housing.

F – Social innovation with environmental benefits* Notes: (*) categories not included in original MEI classification

While acknowledging that generic classifications of eco-innovation types are often used as a signpost for describing general characteristics associated to the innovation process and its framework conditions (e.g. in Kemp, 2011), the literature review and the actual case studies highlighted the need to create tailor-made categories or classifications looking at the entire innovation system. EMInInn looked at the nature of innovation dynamics in the sector, dominating technological regimes (c.f. Kemp and Soete, 1992) and arising environmental and resource pressures. For this purpose each analytical case study pre-defined a minimum set of requirements or characteristics that innovations should posses in order to be considered. As it was noted in the analytical case study of energy (D4.1), the system boundary is often drawn rather narrowly around the technological system or artefact, and wider systemic effects are ignored. Conversely, the analytical approach of this project at the systems level shows the relationship between drawing analytical boundaries and the creation of useful categories, often from known 87


concepts or classifications applied within the industry or policy circles. For example, in the energy case the main features of the energy system and energy innovation were useful to create suitable categories, which lead to the decision to solely focus on technological innovation (and excluding energy using products and energy efficiency innovations). Given the commonly acknowledged fact that internet innovation is not driven by environmental concerns, the Internet case used the 'internet layer’ model, which allowed a timely and suitable classification of innovations and its associated environmental impacts (mostly related to energy efficiency). In the case of transport, the research team decided to create categories based on the environmental performance and framing conditions for innovation diffusion of the different elements of the transport system: car components (catalytic converters, etc.), transport options (green car, bicycles, train), social innovation (product-service systems such as park and ride or bicycle sharing) and system innovations (high speed-train systems). Following the concept of ‘trias energetica’, for the built environment and construction it a useful and resourceful way of categorising innovations in the built environment and construction. However, a number of these innovations clearly relate to the area of renewable energy. The waste case used the well-known waste hierarchy of the EC Waste regulation for the definition of usable categories of innovations different from the counter factual scenario of ‘landfill’. Table 11 below provides a summary of the different criteria for selecting eco-innovations hitherto introduced, together with tailor-made classifications that take into account sector-level and technology-specific features, known environmental performance/issues of industries/sectors, and dynamics of the interplay between innovations and (supposed) environmental benefits.

88

Environmental Macro Indicators of Innovation Table 11 EMInInn’s analytical work packages: categories of eco-innovation, selection criteria and selected innovations for ex-post and ex-ante analyses Work package WP4- Energy Energy innovation based on attributes of the energy system

• • • •

WP5-ICT

1.

Sub-systems of the internet system, following the layer model of ICT system

2.

WP6- Transport

•

Transport innovations, in view of their environmental performance and framing conditions for innovation diffusion

WP7 - Built environment

3. 4.

• • •

1. 2.

Selection criteria of innovations Characteristics of energy innovation (lock-in and path dependency effects, economies of scale, capital intensiveness, network externalities, low R&D intensity) Characteristics of the energy system (resources used/produced, innovations that affect different prats of the generic energy supply chain) Innovation type, with an exclusive focus on technological innovation Impact of innovations on the environment, in particular environmental pressures associated with energy Known environmental impacts of ICT (first, second and third order effects) Characteristic of the internet system of innovation: structure (sectors), institutions, actors and technologies; with particular focus on institutions Activities within the direct internet economy (layer model): front end, networks, software and users Innovation dynamics at the macro and meso levels (multi-level perspective of internet diffusion) The relative (compared to a relevant alternative) environmental performance of the innovation at a product or technology level (microlevel) The capacity of the innovation to induce substantial organizational changes to transport systems The adequacy of the framing conditions (economic, social, normative, etc.) of the innovation for it to diffuse through the economy The magnitude of the rebound effects induced by the technological characteristics of the innovation, with potential to partially or completely offset the environmental gains due to technological improvements.

Innovation aimed at residential buildings Innovation must relate to trias energetica (reducing energy demand,

Eco-innovation classification Resource innovations Conversion and supply-chain technologies End of pipe, waste disposal and clean up innovations System innovations Energy-consuming product and process innovations Energy demand efficiency innovations Front end Access networks Back end Applications

Transport innovations … framed by favorable conditions.

Transport innovations … framed by unfavourable conditions. Transport innovations … and preliminary evidence (e.g. changes in consumption factors) of notable impact of rebound effects. Transport innovations …that have been responsible or have the potential for substantial organisational changes to transport systems. Reduction of the demand for energy

Selected cases for ex-post assessment • Wind energy • Solar PV • Combined cycle gas turbines (CCGT) • Flue-gas desulphurisation (FGD) • •

NA NA

• • • • • • •

NA Mobile phone Smart phone Wireless networks Data centres Electronic product re-use Long distance tourism

• • • • • • • • •

Catalytic converters Diesel engines* Direct fuel injection (DFI) systems* High speed rail systems Battery electric vehicles (BEV) Fuel cell vehicles (FCV) Diesel engines* DFI systems* High speed rail systems

• • •

Park-and-ride (P+R) facilities Car sharing Public bike schemes

•

Insulation of floor, wall and roof / loft

89


Energy efficient residential buildings (housing) based on the concept of ‘trias energetica’

3. 4.

WP8 – Waste

1. 2.

Waste management according to the EU ‘waste hierarchy’

3. 4.

producing alternative energy) Innovation applicable in renovation projects Data availability allowing energy use and direct energy use (availability of sales data in various EU countries, data available on energy savings of innovation as well as embodied energy use; long history so that development could be mapped, around 10-20 years). Interrelatedness and convergence of innovation in the waste system Scale of diffusion of different levels of eco-innovation in waste, namely: (i) prevention innovation; (ii) micro-invention, e.g. landfill; (iii) microinnovation; (iv) management-option innovation, shift of management of a waste type (e.g. municipal solid waste) from a technology option (e.g. landfill) to another (e.g. recycling); (v) organisational innovation (e.g. induced by WEEE or ELV). The expected environmental pressure of waste within a specific waste management option (focus on option iv above). Framework conditions in the waste innovation system (policy, market, actors, relationships, feedbacks, etc.).

Sustainable and renewable sources of energy Most efficient use non-renewable energy (energy efficiency)

• • • • •

Social innovation Prevention Reuse, recycling

• • •

Incineration and energy recovery

•

Landfill

•

Multiple-glazed windows District heating Ground source heat pumps High efficiency (eco) boilers Energy saving light bulbs / LED’s / Fluorescent lights Co-housing NA Material recycling and composting Incineration and energy recovery Landfill

Notes: NA= category of innovation not considered or assessed in the analytical work packages. Source: elaborated based on data and information in Deliverables D4.1, D5.2, D6.1, D71, D8.3 and D9.2 of EMinInn

90


From Table 11 it can be noted that taking into account the functional characteristics of the system where eco-innovations are diffused can have a meaningful explanatory purpose and allows creating a suitable and sound characterisation and classifications of innovation. Such approach subsequently adds value to the empirical findings and interpretation of results. As noted in section 3, taking into account a functions provided/performed by innovations at the system level is one of the characteristics of the sustainability transitions literature, e.g. the provision of electricity, housing, mobility, wireless communication or waste management. An overview of the effects of drivers and barriers of the 26 innovations presumed to be ecoinnovations is presented in Table 12. It is important to note that the following sections on drivers/barriers refer to external factors in the ex-post analysis of EMInInn. The information presented includes: the role of government and institutions (public policy, regulations and standards) and markets (market demand, consumer preferences, price signals). In the subsequent section the topic of synergies and complementarities between barriers/drivers is addressed.

91


Table 12 An overview of the effects of drivers and barriers for (eco)innovation, as identified in the EMInnInn analytical case studies. (Eco)Innovation Wind energy

Flue-gas desulphurisation (FGD)

Perceived effects of drivers or barriers Government and institutions Driver: temporary nature of policy support (public support is provided until technological maturity is reached); medium/long-term policy support (public good nature of benefits of innovation; plans for strategic development of industry; phased reductions over time; subsidies for testing and experimentation); network formation and facilitation of user-producer interactions (e.g. subsidies to early users and producers of turbines); legitimacy (ownership structures allowed direct participation of stakeholders; international perception of environmental accidents and acid rain). Barrier: harmful subsidies (e.g. those given to less environmentally forms of electricity); policy risk associated with measures to overcome environmental externalities; resistance to change (countries with emerging innovation systems and no coherent political voice). Driver: public good of innovation, anticipation to future regulations; political pressure and choices (e.g. Royal pressure to install the best-available technology not entailing excessive costs BATNEEC, favouring energy independence from coal in the US); political elections strategies (considering electorally sensitive regions/states); stringent legislation and regulation (e.g. Clean Air Act and regulation of sulphur abatement stated in the amendment from 1977 in the USA, FGD requirements in Germany, European Large Combustion Plant Directive); testing and certification (e.g. Tennessee Valley authority); standard setting (effectively ruling out competing technologies such as coal washing and lower-

Markets Driver: market creation accompanied by gradual scalingup and learning by-doing and a diversity of business models (successful balance capital costs supported with feed-in tariffs); (modest) cost reductions (achieved by experimentation and learning in test facilities in countries with explicit prioritisation of this industry, such as Denmark); major cost reductions due to increases in technical efficiency of turbines; perception of wind as an export oriented industry.

Barrier: market failures associated with environmental externalities of electricity production; high costs (and suboptimal performance) of early designs, oil cycle (in periods of price loss during an oil crisis it is common that politician give attention to renewable energy, reduced private investments). Driver: anticipation of future scale of production (economies of scale), cost of low-sulphur coal in the US (until the deregulation of the US rail system); market structure (regulated regional monopolies) allowing interiorisation of abatement costs (passed on to consumers); reduction in cost of regulation of SO2 emissions.

Others Driver: geographical factors and resource availability (e.g. electricity needs in rural areas favoured thee development of small scale wind turbines in early 1900s).

Barrier: perceived benefits of centralised electricity grid.

Driver: knowledge diffusion and network creation (e.g. SO2 symposiums), technological learning in successive generations of FGD leading to cost reduction; innovation in monitoring technologies effects on widespread adoption of the regulation.

92


sulphur coals).

Combined cycle gas turbines (CCGT)

Front end (Mobile phone/ smart phone)

Access networks (wireless networks)

Backend (Data centres)

Barrier: (early) government perception of acceptable solutions for amending local air pollution problems (tall stacks and out-of-town locations); resistance to change and lobbying (e.g. direct opposition to proposed EU regulation from industry associations). Driver: legislation promoting co-generation (e.g. Public Utilities Regulatory Policy Act); repealing of legislation restricting gas use (in the late 1980s1990); regulatory pressure to reduce SO2 emissions.

Barrier: high operation costs (of early power plants); relative high costs of (early) FDG.

Barrier: (initial) lack of technical knowledge of broader environmental impacts.

Driver: (niche) markets of gas turbines, increased perception of abundance of resource and falling prices (natural gas); (increasingly optimal) capital and operating costs; risk aversion from private investors to nuclear and coal, and lower risk perception in small-scale gas turbines (compared to coal and nuclear); liberalisation of the (UK) energy market.

Driver: spill-over effects of existing knowledge in a new application area (the knowledge base of underlying technology – gas turbines, jet engines in aircraft, benefitted from R&D support and environmental regulation); higher technical efficiency (compared to open cycle or gas turbines); public environmental and safety concerns of nuclear and coal.

Barrier: legislation restricting the use of natural gas in power generation (e.g. Power Plant and Industrial Fuel Use Act of 1978 and the European Directive of 13 February 1975 75/404/EEC); resistance and lobbying from the coal industry. Driver: regulations and recycling targets (WEEE, California AB2901; RoHS).

Barrier: oil shocks caused rising prices of gas (in the 1970s)

Barrier: NA.

Driver: economic viability of existing collection/reprocessing schemes.

Driver: NA.

Barrier: insufficient end-of-life regulations (reuse, recovery); focus of existing regulation focusing on product groups (e.g. WEEE does not consider individual product types). Driver: energy efficiency regulation, voluntary private sector initiative (e.g. energy efficient data centres and networks).

Barrier: actual recovery rates; insufficient economies of scale (mobiles phones are a minor fraction of category 3).

Barrier: recycling (industrial) processes not including disassembling and separation procedures; design of devices not considering end-of-life

Driver: cost reduction strategies of network operators.

Driver: environmental awareness of ICT-related problems (e.g. resource consumption); CO2 emissions from ICT.

Barrier: NA. Driver: eco-labels (energy star, EU eco-label, Blue angel, 80-plus) voluntary private sector initiative (e.g. energy efficient data centres and networks). Barrier: the lack of standards for monitoring and measuring the energy efficiency and environmental

Barrier: costs of energy from non-renewable sources (compared to renewable sources). Driver: power and operation costs of data centres.

Barrier: costs of energy from non-renewable sources (compared to renewable sources).

Barrier: NA. Driver: advocacy and campaigns from nongovernmental organisations (e.g. Greenpeace) Barrier: (secrecy and) technical design and customation of data centres, which complicates the

93


impact of ICT. Applications (electronic product re-use; long distance tourism)

Driver: product-policy and regulation (EU ecodesign directive);

Driver: cost efficiency (lower price, cost savings) and convenience of e-commerce; growing market of secondhand commerce and virtual auctions.

Barrier: gaps in tax regulations for online private transactions; difficulties to obtain official data from private sector in terms donations and second-hand transactions.

Barrier: NA.

standardisation of monitoring/measurement of efficiency. Driver: eco-design trends for electronic equipment; changing social perception of commerce/auctions of second-hand goods; new patterns in product ownership and quality of life. Barrier:; pace of innovation; design for obsolescence; increasing complexity (and material composition) of ICT technology, which may hamper preparation for re-use.

94


Catalytic converters

Diesel engines

Direct fuel injection (DFI) systems

Battery electric vehicles (BEV)

Fuel cell vehicles (FCV)

Driver: regulation and standards (e.g. EPA regulations on automobile exhaust emissions, European directive 88/76/EEC, Euro 1 standard) and national incentive schemes.

Driver: NA.

Driver: NA.

Barrier: resistance to implement European regulations at Member State level. Driver: fiscal policies; regulations and standards (e.g.. Regulation 715/2007).

Barrier: NA.

Barrier: NA.

Driver: NA.

Driver: highest thermal efficiency, lower fuel consumption and relative lower emissions (compared to petrol engines).

Barrier: (early) asymmetries in domestic fiscal policies on fuels, tax disparities and vehicle registration polices across EU countries. Driver: regulation and standards (air emission standards, e.g.. Regulation 715/2007).

Barrier: NA.

Barrier: NA.

Driver: fuel economy improvements to diesel engines.

Driver: optimisation of the combustion process, leading to higher economic efficiency and lower environmental emissions.

Barrier: NA. Driver: economic incentives (e.g. subsidies, fiscal tax advantages, waiver of road and luxury taxes); procurement policies (e.g. Dutch environmental investment deduction MIA).

Barrier: NA. Driver: NA.

Barrier: NA. Driver: NA.

Barrier: lack of standards (for charging stations).

Barrier: high purchasing prices.

Driver: public support (demonstration projects, R&D subsidies);

Driver: NA.

Barrier: insufficient infrastructure (lack of charging points); low travel ranges of available technology. Driver: NA.

Barrier: NA.

High speed rail systems

Barrier: no commercial fuel cell car in the market; low market acceptance;

Driver: normative framework (Directive 96/48/EC for the inter-operability of the European high-speed rail sysem); national policies for the promotion of large innovation projects (e.g. grand projects d’innovation in France, developing the TGV; Spanish PEIT plan).

Driver: consumer preferences (train over car for European travel);

Barrier: NA.

Barrier: economic efficiency of Countries.

Barrier: no dominant design – technology at the prototype stage; low competitive efficiency of cells; no infrastructure; source of the electricity needed for the production of hydrogen. Driver: growing high-speed rail infrastructure; saturation of existing road infrastructure; existing (potentially upgradable) rail infrastructure; transnational coordination of infrastructure and service. Barrier: NA.

95


Park-and-ride (P+R) facilities

Car sharing

Public bike schemes

Insulation of floor, wall and roof / loft

Multiple-glazed windows

Driver: public policies (parking policies); political will (to implement P+R facilities)

Barrier: (legal) definition of what constitutes a P+D (in relation to traditional parking sites). Driver: information instruments (public campaigns for promotion of alternative transport).

Driver: reduction of car travel demand in urban areas.

Driver: public concerns around traffic congestion and environmental quality; geography and availability of land; public transport preferences; availability of public transport systems.

Barrier: NA.

Barrier: insufficient infrastructure, definitional issues of what counts as P+D

Driver: fee structure (modifies travel planning of user); cost saving (communal) strategies; new business models and niche markets.

Driver: public concerns around traffic congestion and environmental quality; perception of private car ownership; compatibility with greener transport options; distance to travel and flexibility of users; ; vehicle fleet with overall better environmental performance (compared to private cars). Barrier: NA.

Barrier: NA. Driver: public schemes and political willigness (e.g. Danish Bycyke or Paris Velib).

Barrier: NA. Driver: lower cost compared to other transport systems; new business models.

Barrier: NA. Driver: energy efficiency schemes (grants, loans; national insulation incentive scheme), energy efficiency regulation, long term policy planning (e.g. dedicated Ministerial function/Ministry since mid 1970s), governmental campaigns of energy conservation.

Barrier: NA. Driver: increased energy prices (of oil).

Barrier: NA. Driver: social and political concerns over the environment and climate change.

Barrier: fragmented (national) policy approach, short-term (national) policy objectives, (national) policy schemes of short duration, insufficient coordination (national level) of the implementation level (municipalities) Driver: EC policy (energy performance building directive EPBD, including energy performance

Barrier: no natural demand or market for insulation.

Barriers: lack of involvement of other stakeholders (industry, NGOs)

Driver: purchasing power of countries, higher demand for high-added value glass (derived from stricter energy

Driver: societal awareness for energy efficiency and the environment, social and political prioritization

Driver: experience from early adoption; community and non-for profit-run programmes; infrastructure (cycling lanes, docking stations; bike racks and parking facilities); ; ICT and navigation systems; compatibility with other public transport systems; weather and local conditions.

96


measurement method, minimum standards, certification schemes and regular inspection of devices), international agreements on emission targets (Kyoto protocol).

District heating

Barrier: NA. Driver: national policies for substitution of nonrenewable energy sources (tax reform, high taxes to oil, tradable renewable energy certificates, setting/monitoring of energy prices and promoting use of residual heat), subsidies and demonstration projects, network formation for renewable energy production (e.g. Flemish-Dutch ‘heat’ network), inter-operability of standards (combined heat and power and district heating). Barrier: privatisation policies.

Ground source heat pumps (GSHP)

High efficiency (eco)

Drivers: national legislation (tax reduction, tax credits, carbon tax, minimum standards requiring a minimum COP ratio), differentiated subsidies schemes (new/renovation, residential/nonresidential), energy efficiency regulations, interoperability of standards (GSHP use with existing system – floor/wall heating and low-temperature radiators).

efficiency legislation); economic cycle of the construction industry (periods of ‘booming’ markets and increased demand leading to plausible reduction in production capacity and increased prices per unit of glass). Barrier: no natural demand or market for multiple-glazed windows. Drivers: first mover advantages and market formation (early systems installed in the first half of XX century); reliable supply and competitive prices; consumer acceptance, positive perception of (ownership and use of) communal systems.

Barriers: lack of consumer trust and risk aversion; lower price of non-renewable energy sources and abundant supply of fossil fuels; consumer perception of costs of energy source; availability of capital (e.g. equity investments); consolidation of the industry (mergers and acquisitions) and globalisation of energy companies affecting investment decisions (large-scale investments are favoured); extensive price regulation for lower tier costumers. Drivers: private companies efforts to create market for heat pumps via large-scale promotional campaigns (EDF, RWE), cost differential between old (brine-water) and new devices (e.g. new generation of air-to-air GSHP).

Barrier: NA.

Barrier: cheaper cost of sub-optimal technology is detrimental to diffusion of best-available technology (e.g. in terms of efficiency, lower COP for air heat).

Driver: energy efficiency labels, national regulations

Driver: NA

of resource scarcity, involvement of social actors and NGOs (e.g. Friends of Earth).

Barrier: NA. Driver: geographical endowment of energy sources (e.g. natural heat sources in Scandinavia), use for increasing storage capacity (e.g. excess of wind storage).

Barrier: insufficient infrastructure.

Driver: CO2 mitigation potential, geographical demand for heating (northern Europe), increased technical performance (water-heat average COP is 2.5, when optimal is 3.0), increased experience of technicians performing installations (positive learning curve).

Barrier: infrastructural requirements of alternative technology is detrimental to diffusion of bestavailable technology (e.g. air pumps do not require well drilling) Driver: improved energy efficiency of condensing

97


boilers

Energy saving light bulbs / LED’s / Fluorescent lights

for buildings. Barrier: NA. Driver: EU energy label for light bulbs (late 1990s), anticipation to regulatory changes, subsidies, EU product banning, public-private voluntary agreements (e.g. Phillips-EU light bulb manufacturers-Government phase-out plans; retailers-energy companies agreement in the UK to phase out light bulbs 9 months ahead of EU banning), formation of networks (e.g. Dutch Taskforce for Lightening), explicit role of lighting for CO2 abatement in national energy policy (UK’s Climate Change Act, CERT programme). Barrier: NA.

Co-housing

Waste management technologies

Driver: explicit design standards (eco-villages, energy efficiency in buildings, passive house), public policies (subsidies), soft-institutional factors: sustainable consumption habits, (shared) community values, desire to belonging to a (sustainable) community, ‘de-growth’ and ‘sustainable lifestyles’ movements. Barrier: definitional issue in policy schemes of what constitute a co-housing (vs. collective housing). Driver: long term EU policy/regulation - general and specific binding legislation on waste streams (Waste directive); economic instruments; performance targets (environmental management); technical regulations (ELV, WEEE, RoHS); promotion of industrial networks for waste recovery/recycling; complementarity between environmental policy and innovation.

boilers over traditional product (traditional boilers). Barrier: NA. Driver: affordable price of new lighting alternatives, private investments (e.g. in campaigns), sufficient sales volume and market creation (derived from the influence of public intervention to ban the old product).

Barrier: initial no- natural demand or market for energy efficiency lighting, mislead perception of new product (e.g. LED) as having no clear advantages for users. Driver: social perception of efficiency gains from shared living, cost savings.

Barrier: NA. Driver: product obsolescence in related industries (e.g. electronic waste and cars) increasing recycling quotas; recycling markets and cost-benefit considerations of ‘waste’ generator industries (e.g. cars, electronics, etc); cost of raw materials; evolution of the value chains in the different waste/recycling/recovery sectors;

Barrier: policy myopia (when outcome of policies in achieving policy targets/objectives are too slow). Barrier: NA. Note: NA: empirical evidence is not available within the corresponding EMInInn report

Barrier: NA. Driver: industry-coalition lobbying for phase out of old technology (incandescent bulbs), increased environmental awareness from Philips-sponsored film ‘an inconvenient truth’.

Barrier: NA. Driver: environmental awareness (e.g. environmental footprint/person of co-housing is half of that of an individual ‘regular’ house), age of target user (more common in >50 years old people), inter-operability with other social innovations (e.g. car sharing, sharing of eco-travel equipment, shared allotments). Barrier: individual consumption, lesser social and community ties of (young) professionals). Driver: complementarity between technological and organisational innovation; geographical factors.

Barrier: NA.

98

FINAL 15/04/8 Derived from the above, the remainder of this section addresses some of these major drivers and barriers to eco-innovation, as identified in the analytical work of EMInInn. The following subsections provide a number of salient messages from each analytical work package. However, such overview is not a comprehensive review of the information contained in the corresponding deliverables (D4.1, D5.1, D6.1, D7.1 and D8.1). The remainder of the section has been written as a illustration of the different factors and conditions under which innovation diffusion processes generate environmental benefits. 7.1.3

The role of government and institutions: public policy, regulations and standards

As it can be noted in the summary table presented in the previous section, all analytical cases show the important role of government in enabling innovation. This positive influence is often enabled by setting up standards, regulations and other means of public intervention such as promoting learning and experimentation, building networks or phasing out (less efficient) products. In a number of cases, government agencies, standardisation and verification bodies and public support to R&D played an important role in allowing the technologies to reach its optimal development, hence facilitating its diffusion. However, the role played by government differed substantially—with CCGT technologies, government’s role was largely early R&D and the subsequent removal of barriers to diffusion (through deregulation of energy markets and the removal of prohibitions on the use of gas for power generation), rather than through sustained public deployment support. In the case of bicycle sharing schemes, countries building sufficient infrastructure allows them nowadays to enjoy the benefits of higher diffusion rates, and (implicitly) a more positive contribution of the innovation to their environment. Below some key messages from each of the five analytical case studies are briefly discussed. The energy case study provides important lessons in relation to the political dynamics between innovation, regulation and environmental performance. Flue-gas desulphurisation (FGD), an end-ofpipe technology, emerged as a direct response to a regulatory requirement (at Battersea Power station 1931). The case study also highlights the importance of expectations concerning regulation (expectations induced significant innovative activity in the 1960s and 70s). Equally illustrative are the effects of other institutional factors for the diffusion of end-of-pipe technologies, such as the setting of standards, verification of performance and the promotion of best-available technology not entailing excessive costs (BATNEEC) – topics that remain of high importance in modern EU environmental policy. Political dynamics have been very important in shaping the incentives for this technology from the inception: for example, the development of FGD in the US in the 1960s and 1970s was politically preferable to the use of low-sulphur coals, because the high-sulphur coal regions in the Eastern US were electorally important. The development of FGD has enabled reductions in sulphur emissions, but as at Battersea in the 1930s, the existence of FGD has enabled continued use of coal that might otherwise have been shut down or never built. One can thus identify a feedback through the regulatory system in which the existence of the end-of-pipe innovation allows continued use of the (still partially) polluting activity. At the same time, the successful development of FGD reduced the costs of regulating SO2 emissions, and arguably enabled much stronger regulations than would otherwise have occurred. A second, and probably more powerful, regulatory feedback is thus observed in which end-of-pipe innovation induces (or at the

99

FINAL 15/04/8 very least facilitates) more stringent regulation. While it has been common to argue that regulation induces innovation, the role of innovation in enabling regulation is less commonly articulated. For the internet analytical case, an example of an enabling technology, it is generally agreed that the environment has not been one of the principal drivers for innovation in the sector. This case uses the multi-level perspective of sustainability transitions (c.f. Geels, 2002) to review general drivers and barriers of eco-innovation at the landscape and regime levels. In this review, the research team notices the existence of a number of external pressures that may result in either relative or absolute ‘greening’ of the internet system of innovation, particularly in terms of energy efficiency. In this regard, a variety of policy initiatives (particularly regulatory approaches to electronic waste and to eco-design, such as the WEEE and RoHS directives), and market-based pressures have been instrumental in the objectives of achieving a better environmental performance. When analysing particular technologies, (end-of-life and energy efficiency) regulation, information tools (eco-labels) and voluntary initiatives appear as major drivers for the diffusion of internet innovations. For access networks and backend, voluntary private sector initiative seem to been acting as drivers of innovation, for example companies listed as hosting energy efficient data centres and networks (e.g. the Dutch cleanbits initiative and the Green Web Foundation) and the growing amount of eco-labels focusing on electronic equipment and backend-equipment in general (for example the energy star, EU eco-label, Blue angel, 80-plus). Among barriers under this category, the lack of standards for monitoring and measuring the energy efficiency and environmental impact of ICT is probably one of the major factors to be cited. For front end and applications, the design and focus of existing regulation focusing on levels of aggregation that inhibit recycling and/or reuse are highlighted as hampering factors to innovation diffusion. For example, the WEEE directive does not consider individual product types but on equipment categories, hence recycling targets are difficult to be achieved due to complications in the separation and recovery of metals and plastics. The transport case makes evident that promoting markets and creating lead users via subsidies and other tax incentives could well not be the most effective way to also achieve positive environmental gains when the overall transport system is not considered. This case study shows that, given their very low diffusion rate, that promotion of FBE and HFC vehicles through subsidies (being the dominant policy approach) can take away a good part of the environmental benefit from using those vehicles. The solution to sustainable transport probably lies more in (higher) product taxes for gasoline vehicles and diesel vehicles and other environmental harmful products than in subsidies for FBE and HFC vehicles. In contrast, public incentives in alternative transport modes (e.g. bicycle sharing systems), promotion of system-level solutions (e.g. high speed train and park and ride schemes) and development of appropriate infrastructure (e.g. for bicycle sharing schemes, charging points for electric vehicles) are key to their wider diffusion and overall positive environmental benefits. For the built environment and construction analytical case study, one of the most important drivers for eco-innovation diffusion identified is long term governmental policies. The effects of strong governmental regulations (energy performance of buildings), voluntary agreements and product bans / phasing out harmful products (e.g. industry-led substitution and subsequent prohibition of incandescent lighting across the EU) can provide positive influence on the overall environmental benefits of this sector. This analytical case also sheds some light on National differences (and local 100

FINAL 15/04/8 characteristics) and the perceived effects of government intervention as a driver to the diffusion of eco-innovation. The case of GSHP shows how Sweden and Austria presented high diffusion rates due to a strong policy to support heat pumps by means of a tax reduction in the former and strict energy efficiency regulations for the latter. The cases of insulation/multiple glazing, high efficiency boilers and energy saving lighting shows that stricter compliance and long term planning in the UK (compared to the Netherlands) was an important factor for diffusion at (faster and) the larger scale. The energy efficiency lighting case showed how anticipation to regulation (e.g. EU product ban of incandescent light bulbs) can be a stimuli for industry to develop coalitions and to promote a swift product substitution to a more efficient (and of higher-added value) option. Finally, the case of cohousing shows how soft-institutional factors such as values, environmental awareness, and consumption habits are strong driving factors favouring a collective form of housing, which in turn promote more other social innovations, especially ‘shared-based’ schemes. The most evident barrier of most (eco)innovations considered in the built environment and construction case is related to the (apparent) lack of a natural market for these technologies, which is further discussed in the following section. Most of the detailed work on the regulatory drivers of eco-innovation has been undertaken in the waste case study using econometric analysis. 65 Overall, the empirical work in this work package found that the shift is a firm priority of EU waste policies (in particular the EU ‘waste hierarchy) and these policies have been a major driver of the eco-innovation diffusion. The main factors to be considered are, inter alia: (i) general and specific binding legislation on waste streams; (ii) economic instruments, like landfill taxation; (iii) targets on management change to be achieved by defined deadlines; (iv) technical regulations; (v) the promotion of industrial networks for waste recovery/recycling (e.g. packaging). Key points from the waste case studies around the role of regulation promoting eco-invention are further specified below: • Regulation does stimulate invention (in waste technologies), as well as stimulating diversion from landfill (i.e. increase recycling and energy recovery).. However, policies designed to prevent landfill have not had any effect on waste generation (D8.3). • The effect of policies in achieving policy targets/objectives can be slow and sluggish because of the long time required for the diffusion of alternative technologies of waste management with respect to status quo (D8.3). • Organisational innovation can be an important outcome of policies, as in the case of private/public compliance scheme foe WEEE developed as a consequence of EU policy provisions, and it can spur also technological innovation while possibly orienting the latter, e.g. innovation diffusion in recycling instead of innovation in product making (material mix) and prevention (D8.1). As noted earlier, the waste management case unveiled that the macro-level eco-innovation has been meant as the observed adoption/diffusion of innovations in waste management/recycling that reduce environmental pressures compared to landfill, which represents the baseline in terms of pressures. A significant, although gradual and smooth, shift from landfill to other waste management technologies (incineration and recycling) has been observed in the EU during the last 20 years. 65

This analytical case study used eco-patent data as a proxy to eco-innovation invention; hence it is important to take this into account when interpreting the results.

101

FINAL 15/04/8 7.1.4

Market, consumer preferences and price signals

Similar to the overview of policy drivers/barriers presented in the previous section, all analytical cases show the important role of markets, price signals and a number of consumer-related aspects in enabling the diffusion of innovation. Consistent with the broader literature, economic factors, market demand and customer requirements can be identified as drivers of supposed ecoinnovations. In the analytical case of energy most barriers were of economic (and technological) nature, especially during the initial stages of the development of the technology and the early formation of the innovation system. For the case of wind energy, market failures associated with environmental externalities of electricity production; high costs (and suboptimal performance) of early designs, oil cycle (in periods of price loss during an oil crisis it is common that politician give attention to renewable energy, reduced private investments). For the case of FGD it was the high operation costs of early power plants and relative high costs of early FDG that played a role for slowing down its diffusion. In the case of CCGT the aversion to invest in competing technologies (coal and nuclear) due to safety, environmental and the required large scale investments, and decreasing costs of the energy source (natural gas) facilitated a more positive perception of investors, since investing in smaller scale power plants using gas turbines was considered feasible. Once these barriers were overcome the pace of innovation diffusion was relatively straightforward. But again, markets did not do all the job – as it is shown in the following section it was the combination of public policy that created favourable more favourable conditions for (a faster) diffusion of these technologies. For the case of internet innovation, due the global nature of the internet system, the empirical findings highlighted a number of implications for pollution and emissions leakage to locations with weaker environmental standards; and the market failures associated with most environmental pressures (i.e. damages are externalities not reflected in market prices in the absence of policy). In addition, it is observed that the pace of innovation in the sector creates challenges for governance and regulation while also fostering rapid product obsolescence. For example, for access networks and data centres, costs is the sole and major driver for an eco-efficient design and operation of this type of artefacts. Being electricity costs up to 50% of the purchase price of a server over its life time, energy efficiency is an important factor in the resulting power and operation costs. For the case of mobile devices, it is consumer preferences (demanding larger battery durability) a key factor driving energy efficiency of devices. For the case of applications, such as the reuse of electronic products, it is convenience and the reduction of transaction costs. In terms of barriers, and due to the close link of energy efficiency and waste management of ICT, identified market barriers are related to price signals (from the non-renewable energy market) and insufficient recovery rates of e-waste, which contributes to a sub-optimal cost-structure / economies of scale. The work of EMInInn on transport innovation, in particular when referring to alternative transport, highlights the role of consumer preferences, social perception of environmental problems, changes in the perception of ownership and community, travel demand and congestion, and cost of transport systems. In the case of the built environment and construction, in the case of district heating it was consumer-related factors that hampered diffusion in countries with little cultural and social preferences for system-level provision of energy. The lack of consumer trust and risk aversion; lower price of non-renewable energy sources and abundant supply of fossil fuels; consumer perception of 102

FINAL 15/04/8 costs of energy source in the Netherlands were detrimental factors for a wider diffusion of the innovation. In terms of economic and market factors, lesser availability of capital (e.g. equity investments), the process of consolidation of the industry (via mergers and acquisitions) and the globalisation of energy companies affecting investment decisions where large-scale investments are favoured (and not local-level investments, such as those required by district heating companies0); extensive price regulation for lower tier costumers. In the case of GSHP it was the initial cheaper cost of sub-optimal technology has been detrimental to diffusion of the best-available technology (e.g. in terms of efficiency, lower COP for air heat). Perhaps one of the most the most revealing pieces of evidence of the role of markets is that of no natural demand or market for multiple-glazed windows, insulation and energy efficient lighting. For the case of the latter, it was also affected by the (perhaps mislead) perception of the clear environmental and saving advantages of the new product (e.g. LED). Similar to other cases discussed above, market formation was only successful because of the intervention of governments and large companies to do it so. For the waste case, while waste innovations (invention, diffusion) are expected to be mainly policy driven, market variables (e.g. virgin material prices) and the evolution of the value chains in the different waste/recycling/recovery sectors, albeit little documented, are relevant for technology invention/diffusion. Hampering factors for technology diffusion is attributable to capital investment needs; treatment capacity changes by jumps at the local level, smoothly at the macro level. Microeconomic considerations (by public administrations and private companies) can be very relevant for investments able to shift management towards different technologies. Finally, markets for recyclable materials can be affected by imperfect information, and consequently market failure (for instance, the presence of contaminants in used waste oils, the structural strength of scrap, the mix of plastics in a given bundle, etc.). 7.1.5

Interaction and synergies between drivers and barriers (inc. complementarity)

In section 3 it was noted the intrinsic nature of drivers and barriers to innovation. It was then noted that while barriers and drivers may exert synergistic or opposite effect along different stages of the innovation cycle, depending on changing conditions in markets, sectors or technology. The results of the energy analytical case clearly illustrate such intrinsic and synergistic effect of drivers and barriers, partly condition by sector-level dynamics of innovation in the energy system. In general, it is observed that there must be market opportunity, created by policy, created by expectations about future policy, created by direct social and political pressure on firms, and created by the existence of markets for electricity in general and clean electricity in particular. At the same time, there must be technological opportunity, both in terms of the potential of the technologies in question, and in terms of the availability of resources required for their effective use. The example of CCGT, a non-intentional eco-innovation, is highly illustrative of the above. This technology emerged following the withdrawal of specific barriers, such as policy restrictions on the use of natural gas as a power generation fuel, and the enabling of new entrants into existing power markets. Public funding did play an important role in the development of the underlying gas turbine technologies, but this was not an innovation that required dedicated and long-term public funding. For the internet case, markets and consumer preferences (and choices) constitute the major driver of innovation and gains in energy efficiency. Relative increases in energy and resource prices, as well 103

FINAL 15/04/8 as the market imperative for hand-held devices to improve functionality given weight, size and battery restrictions, were signposted as drivers of relative environmental improvements. Findings of the built environment and construction analytical work package further provide insights on the historical interaction between the drivers and barriers. For example, the effect of some ecoinnovations (no incandescent light), can lead to shifting consumption (led and other energy saving lighting) production chains. In other cases, the positive effects of energy efficiency measures (innovations) are to a great extent being offset by consumer behaviour (rebound effects), due to an increase in service and comfort level. For the waste case, the empirical findings suggest an apparent complementarity between innovation and environmental policies for determining levels of recycling. A notable finding of the empirical work of this case study is that market forces are not sufficient, given the relevant 'public good' nature of such innovations. The results of EMInInn show that both ‘regional market forces’ – agglomeration – and usual geographical factors are not significant, while some ‘policy contents’ of the regional framework are. Moreover, the higher rates of innovation in the waste sector may be associated with more-rapid product obsolescence in connected industries (e.g. car industry), resulting in larger volumes of waste from end-of-life (vehicles). Note that the degree to which the associated negative environmental pressures have been offset by a more-rapid adoption of more-efficient vehicle technologies has not been estimated.

7.2 Framework conditions The literature reports that the perceived effect of them tend to vary depending on the type of innovation, the size of the firm, the sector, the geographical location, the business cycle and the stage of development of the (national or technological) innovation system. The factors hitherto cited constitute the so-called framework conditions. As noted in section 3, a systemic approach to innovation requires providing a balanced account of the conditions surrounding innovation. The innovation systems literature is particularly resourceful in analysing how well-functioning innovation systems constitute a pre-condition for innovation diffusion. In the remainder of this section the outcome of EMInInn’s work on framework conditions is briefly presented. 7.2.1 System and market failures As noted in section 3, the specialised literature notes that barriers to eco-innovation may be understood as stemming from market failures and system failures acting in them. Systemic failures arise when different activities in in the enabling environment (innovation system) are not conducive for eco-innovation e.g. not sufficient national R&D capacity, not enough applied research in emerging knowledge areas. In this section we briefly sketch some findings from the analytical work of EMInnInn that shed light on system and market failures and how these affect the innovation system, hence hampering or facilitating the diffusion of presumed eco-innovations. As a matter of providing an example of market and system failures, the analytical case of the built environment identified market and system failures as part of a broader analysis of innovation . Using a systemic failure framework it was possible to identify, from the general literature, the main system-level failures hampering (general) innovation in the sector. Such analysis provided useful information about failures in terms of: infrastructure and (forma/informal) institutions, capabilities, cooperation (and other interactions) and market failures. A summary of results are presented in the following table.

104

FINAL 15/04/8 Table 13: Particularities of the sector classified by the framework [pages 23-32 of the Deliverable7.1] CONDITIONS infrastructure Lacking energy infrastructure that enables local production and consumption and enables sharing between facilities

SYSTEM FAILURE

formal institutions

informal institutions

ACTIVITIES

Capabilities

• • • • • •

The level of knowledge is limited and is insufficiently exchanged between relevant actors Lack of learning across projects hinders innovation diffusions Lack of management skills for complex project based collaborations Small companies lack skills and resources for R&D, innovation and change Cooperation is hampered by project based coalitions One-off projects discourage innovation diffusion and investment Public private partnership weakly developed

Demand

• • •

No ‘natural’ demand Unclear benefits hinder market development for sustainable innovation Lack of trust between client and contractor

Structure

• • • • • •

Market dominance hinders entry of new actors and solutions Highly cyclical nature of housing sector affects knowledge development Fragmentation and a lack of differentiation Increased market dominance through vertical integration Split incentives hinder investments Unfair competition by fossil as externalities are not priced in

Interactions

MARKET FAILURE

MARKET

Need of new rules and regulation, current regulations are based in the old central energy production paradigm • Building industry is highly regulated, it hampers innovation • Short-term and unreliable governmental policy frustrates investment • Current policies focused more in incremental rather than in radical innovation • Formal tendering processes and subsidies are too prescriptive Informal institutions and consumers have to get used to the new reality, they do not take risk to look for alternatives • Conservative culture dominated by vested interest • Claim culture leads to risk averse behavior • Difficult understanding of different cultures and languages

Externalities /Split incentives

•

The analytical case of energy is exemplar in providing evidence on the need to have well-functioning innovation system – that is, having adequate framework conditions by overcoming barriers and failures in the system. In this regard, it is noted that successful eco-innovation diffusion requires that the innovation system must be functioning effectively – that is, it must be generating knowledge, enabling that knowledge to be diffused through linkages between actors; it must be enabling entrepreneurship, experimentation and learning, and it must be capable of doing the work required to establish legitimacy, develop markets and attract (and absorb) resources. In addition, each of the activities of the innovation system must be oriented towards sustainability. This includes a role for strategic choice and appraisal of which technological opportunities to pursue, and which market opportunities to support. The wind energy example provides useful lessons in this regard, as its diffusion has been enabled by the promotion of entrepreneurship (and a variety of business models), generation of general knowledge (via applied R&D), experimentation (using small-scale testing), guidance of the search (with long-term policies for the development of an export-oriented industry), etc. An aspect often neglected in innovation diffusion studies and sustainability transitions is their spatial and geographical dimensions and how these are interrelated with framework conditions (c.f. 105

FINAL 15/04/8 Coenen et al., 2012). The analytical cases of the built environment and construction and waste provide important lessons in this regard. These cases exemplify how local and regional institutional quality (government and governance) and modes of implementation of public programmes can influence the pace and the depth of innovation diffusion of innovations (e.g. waste management, insulation and multiple glazing). These cases show s how the local innovation environment and innovation networks (universities, research centres, co-housing communities) can influence the propensity to diffuse environmental innovations. Finally, the waste case notes that the general innovativeness of the company can largely influence its attitude in environmental innovation adoption, and different innovations are often correlated in adoption at the level of the company. 7.2.2 Framework conditions and policy design Earlier in this document it has been extensively mentioned that the identification of barriers to ecoinnovation and the identification of market and system failures are key for the design of effective policy interventions. In section 3 it was also noted that more and more, policy design for ecoinnovation requires adequate mixes looking at different stages of the innovation cycle. The evidence produced by the analytical cases of EMInInn can be used to illustrate how to take into account the framework conditions surrounding particular innovations when policies are being designed/reformulated. The energy case again provides useful lessons policy design, in particular when there is a need for balancing incentives for cost reduction with enabling market development and learning-by-doing. In the case of wind power during the formative stages of the technology, cost reductions emerged perhaps counter-intuitively, not in those countries that focused on policy supports designed to yield the cheapest possible deployment of wind power, such as the UK and Sweden (Mitchell and Connor, 2004; Bergek and Jacobsson, 2010). Rather, it was countries that prioritised strategic development of the industry and that provided support with phased reductions over time that enabled firms to develop and innovate to reduce costs (Jacobsson and Lauber, 2006; McDowall et al., 2013b). This case demonstrates that attempts to focus on cost-optimal solutions too early in the presence of the enormous uncertainties of innovation are likely to be mistaken. Fostering promising portfolios of technologies, and addressing potential system failures, can be more effective in the long-term. In terms of policy monitoring, the transport case exemplifies how stringency in the implementation and adoption rate (e.g. catalytic converters in the Netherlands vs Portugal/Spain). Lead countries such as France and the UK are clearly showing the way in the diffusion of sustainable transport modes, but it is important to carefully consider the several trade-offs are observed between life cycle stages and/or environmental indicators when comparing the product-level environmental profiles of innovations with those from their relevant alternatives. Such results invite caution when comparing (alleged) eco-innovations with their alternatives. Rarely do the former offer a better environmental performance in all aspects. The analytical work of EMInInn also sheds light on the complementarity between innovation policy and a number of additional policy areas. In the energy case, many energy and climate change policies are justified partly on the basis of their innovation co-benefits. The review of European, UK and Swedish energy and climate policy processes (see EMInInn Milestone 19) makes clear that governments frequently argue that climate and energy policies will have innovation benefits, in addition to the direct energy or climate benefits. However, this claim is rarely based on any explicit ex ante assessment of the expected innovation benefits of a particular policy. Moreover, few 106

FINAL 15/04/8 examples could be found in which evaluations of policies have attempted to identify the induced innovation benefits attributable to energy and climate policies. Many technology-specific policies will certainly have some effects on innovation. A further question—much discussed in the literature—is the extent to which the type and design of policy instrument, alongside wider supports to the innovation system or the framework conditions of broader environmental policy, might foster or hamper better innovation performance. For example, there has been considerable discussion about the relative merits of feed-in-tariffs vs. tradable certificates in inducing innovation in renewable energy technologies (Bergek and Jacobsson, 2010).

7.2.3

Framework conditions required to realise the full environmental potential of resource-saving eco-innovations As has been discussed extensively in earlier sections of this report, the presence of rebound effects can reduce the anticipated environmental savings associated with innovations that use inputs more efficiently to produce a given unit of goods or services. Taking the example of energy, energy efficiency innovations result in a smaller reduction in demand than might otherwise have been expected from a simple engineering assessment, because the efficiency improvement changes the effective price of the energy service, resulting in a shift towards higher demand. The question then arises: what are the environmental consequences of that induced demand? The environmental consequences of rebound effects depend on the environmental profiles of the goods and services for which demand is induced. This may comprise some mix of both direct rebound (in which demand is induced for the consumption of the energy service that has become more efficient), and indirect and economy-wide rebound, in which demand is induced across a range of goods and services as a result of the re-spending of the money saved as a result of the efficiency innovation. This environmental profile is, of course, changing over time in response to consumer preferences, innovation, and policy signals, among other factors. Importantly, the environmental profile of consumption increases that occur as a result of rebound effects can be influenced by policy. Both regulations and pricing policies will limit the negative environmental consequences of consumption induced by innovations that increase efficiency, as has been argued by van den Bergh and many others (van den Bergh, 2013).

This can be illustrated with the results of an economic equilibrium model. In Figure 22, we show how the carbon intensity of a marginal unit of consumption across the European economy changes as a function of a carbon tax. The figure illustrates that the environmental consequences of a rebound in demand will differ markedly across different carbon prices.

107

FINAL 15/04/8

Increase in emissions (m tCO2e) for a marginal increase in GDP (m€)

GHG intensity of marginal unit of GDP 2.5 2.0 1.5 1.0 0.5 0.0 10

20 30 40 Carbon price (Euros per tonne CO2e)

50

Figure 23. Figure showing carbon intensity of energy consumption in 2030 under different carbon price scenarios. Source: EXIOMOD modelling

The DILER model, explained in section 5.2.2, described an approach to assessing the potential environmental consequences of rebound effects induced by the diffusion of a specific innovation. For ex ante assessment, this could be combined with different carbon price scenarios to examine the sensitivity of ‘environmental rebound effect’ to carbon price.

108

FINAL 15/04/8

8 Summary of insights from case studies and from macro-modelling 8.1 Insights from case studies The case studies in EMInInn provide a set of assessments of a number of key technologies that have been identified as having potential to improve Europe’s environment. A summary of these assessments is provided in the table below. Table 14 Summary of key insights of technologies assessed within EMinInn

Assessment method

Key insights

Wind

Ex post scenarios with IOHA. Simple assessment of indirect rebound effects (income effect only)

Wind can be considered to be an eco-innovation when compared to relevant alternatives, taking into account the full production cycle through IOHA.

PV


PV has generated an ‘environmental debt’ – the pressures generated through the installation of European PV capacity had not yet (by 2010) been offset by the reduced environmental pressures associated with emission-free power generation.

CCGT


CCGT can be considered to be an eco-innovation, though the extent to which rebounds have been assessed is limited.

FGD

Ex post scenarios with IOHA, using two alternative historical counterfactual scenarios.

FGD has generated significant reductions in SO2 emissions to air, though other abatement innovations have also played a significant role. It seems likely that FGD has enabled greater use of coal than would otherwise have taken place, but this depends on assumptions about what political decisions would have been taken if FGD had not been available. This is obviously highly uncertain.

Diesel engines

DILER

Diesel engines are more efficient than petrol engines. However, policy support for diesel engines, through subsidies for diesel fuel, may have generated GHG emissions greater than the counterfactual, in which petrol vehicles remained dominant.

Direct Fuel

DILER

The economic savings from the more efficient DFI 109

FINAL 15/04/8 Injection (DFI)

systems caused a rebound effect that took back an important share of the CO2 emission benefits. However, moderate CO2 emission reductions still took place as this technology diffused.

High speed rail (HSR) systems

DILER

The diffusion of HSR systems caused a backfire effect for the global warming potential (GWP), land use change and abiotic depletion potential indicators, so all environmental gains were taken back by the rebound effect.

Park-andride (P+R)

DILER

The notable increase in costs from the use of P+R facilities translated into a noteworthy negative rebound effect. The bound consumption thus enhanced the GWP emission savings from this innovation.

Car-sharing DILER schemes (CSS)

The diffusion of CSS entailed an increase in GWP emissions due to a backfire effect. This extreme version of the rebound effect stems from notable costs savings and a low environmental intensity with respect to the rest of consumption.

Bicycle sharing schemes (BSS)

DILER

The use of BSS caused a remarkable backfire effect, which stemmed from both notable economic savings and a very low environmental intensity reference.

Insulation measures

Life Cycle Inventory on energy balance, LCA for materials (embodied energy) historical product penetration figures

Insulation measures bear a huge energy savings potential, but may be offset to a great extent by the user behaviour as a consequence of having a more energy efficient dwelling. Domestic energy consumption reductions causes a reduction in GHG emissions.

Double glazing

Life Cycle Inventory on energy balance, LCA for materials (embodied energy) historical product penetration figures

Double glazing has a rather successful penetration level in EU housing stock, reduces heat losses considerably, but real savings potential may be offset by user behaviour.

Energy from waste

LCA with counterfactual historical scenarios. Potential economic feedbacks not

Energy from waste appears to have reduced environmental pressures compared with landfill; the extent of GHG savings is dependent on the counterfactual power generation technology (CCGT or coal).

110

FINAL 15/04/8 assessed.

8.2 Insights from Macro-modelling The macro-modelling undertaken within EMInInn, which made use of three different Computable General Equilibrium models (EXIOMOD, GDynE and a model developed for EMInInn by Adriaan van Zon at MERIT), is described in detail in deliverables 9.2 and 9.3. This section provides an overview of the key insights from an innovation policy perspective. Deliverable 9.2 described the methodological approach for linking micro-level technology data with the CGE model EXIOMOD. The modelling assessed the impacts, in terms of macroeconomic and macro-environmental outcomes, of diffusion of a set of specific innovations. The impacts of six ‘packages’ of specific improvement measures were modelled: (1) Built Environment area, (2) Transport area, (3) Energy area, (4) Waste area, (5) ICT area and (6) all improvement areas combined. The impacts were explored with EXIOMOD model for the period 2015-2050 as compared to the official baseline scenario of the European Commission. Emissions reductions arise from implementation of most packages, with the largest impacts being attributed to the Energy and Built Environment innovations. Transport and Waste improvement areas lead to slight increase in emissions. This is due to construction phase effects in case of Transport and boost in economic activity (increase in recycling) in case of Waste. Note that the modelled innovations do not attempt to represent innovation in general in these sectors, but are instead very specific technologies. It would be a mistake, therefore, to see the results of the transport package in terms of the impacts of innovation ‘in transport’ in general. Implementation of Transport and Built Environment improvement areas leads to negative effects on EU28 GDP. Built Environment improvements are related to less use of electricity services and own production of electricity by households. This leads to lower demand for electricity and hence lower production level. As a result the overall effect on GDP becomes negative. Implementation of Transport improvement area leads to more demand for electricity and less demand for gasoline. The overall impact on the GDP is small and negative. This result is a result of the specific innovations examined, which are rather highcost, and for which key components may be imported rather than produced in the EU, with impacts on the modelled GDP. GDP-level impacts of all considered scenarios are small and point to the fact that implementation of the technical improvement options does not have any significant effect on the overall level of employment or change in the Total Factor Productivity (TFP) of the economic sectors. Impacts of the considered scenarios increase over time that reflects the increase in the penetration rates of the technical improvements. Impacts on the use of water and use of various types of resources differ between the considered scenarios and 111

FINAL 15/04/8

reflect the differences in their construction stage impacts. All considered scenarios (expect for ICT) have positive environmental impacts in terms of emissions taking into account both their construction and use stage effects. This means that the chosen improvement options can indeed be seen as eco-innovations. Rebound effects The modelling in WP9 identified potential rebound effects arising from the diffusion of the resource-saving innovations examined 66. The rebound effect is defined here as the environmental savings that are thwarted or enhanced as a consequence of behavioral and systemic responses to technical changes. Because technology-based or engineering estimates of environmental savings generally fail to capture such responses, the actual savings might differ once these are incorporated. The rebound effect is thus commonly understood as the difference between the engineering and the actual savings. For this we compare two approaches; firstly, a partial equilibrium (PE) approach that provides us with both direct rebounds, and secondly the Computable General Equilibrium (CGE) approach used in Deliverable 9.2, that includes both direct and indirect rebound effects. The exogenous shock we introduce is based on the direct effects of the technical improvements/eco-innovations used in the computable general equilibrium model EXIOMOD used in Deliverable 9.2. We see a clear consumption shift compared to the baseline in the partial equilibrium approach, for both the eco-innovations introduced in the Transport WP and the Built Environment WP. This consumption shift, i.e. a redistribution of consumer expenses, can also be seen as a rebound effect (indirect). So, the eco-innovations have opposite effects on different product groups. Moreover, in case of the eco-innovation introduced in the Built Environment WP, we see that the change in fuels (all energy sources, except electricity) and metals (and minerals, however effect is smaller) use for the CGE is larger than resulting from the PE simulations. Since a CGE model captures wider economic effects, i.e. both direct and indirect effect of the eco-innovations and the PE model only measures the direct effect. For the ecoinnovations analyzed in the Transport WP we see that the change in fuels, biomass, metals, and minerals use for the CGE is larger than resulting from the PE simulations. The relative PE results are very small (around zero, i.e. almost equal to the baseline). However, we must state that the CGE includes the changes in production (in each sector) and the PE only includes changes in production plus the changes in intermediate consumption between sectors caused by the input output linkages. Comparing the emissions output of the PE results with the CGE results, we see that some of the environmental gains (less emissions) resulting from the eco-innovations is lost, i.e. the PE show lower emissions relative to the baseline, than the CGE results. This can also be 66

We only modelled the Partial Equilibrium model for the Transport and the Built Environment WP.

112

FINAL 15/04/8

identified as a clear rebound effects. This also holds for the both the eco-innovations analyzed in the Transport WP and the Built Environment WP. The impacts on considered improvement areas on the exports and imports of EU28 countries are small. The total impacts on exports are almost equal to the total impacts on imports, which is due to the trade balance constraint. Finally, the sensitivity analysis shows small variations in the results. This indicates that the model results are robust. The changes in production emissions are more affected than the changes in GDP. The sensitivity analysis was done by varying the elasticity of substitution between capital and labour by multiplying and dividing it by two. Recommendations for impacts assessment studies •

•

•

•

•

Impacts of individual eco-innovations on GDP and other macro-economic indicators are quite small which means that it makes sense to assess packages of improvement options with macro-economic models It is important to take a systemic view on the impacts of eco-innovations. For example our analysis has shown that environmental impacts of electric vehicles strongly depend upon the electricity mix of EU countries. The combination of renewable electricity with introduction of electric vehicles has the highest environmental benefits. Results of ex-ante analysis have the highest sensitivity towards the assumptions about the penetration rates of eco-innovations. This means that sound scientific research is needed to be able to provide such forecasts. Impact of rebound effects on the environmental effects of eco-innovations strongly depends upon the type of innovation. The highest rebound effects are associated with overall improvement in energy efficiency. Proper sensitivity analysis is an important part of ex-ante impact assessment of ecoinnovations

113

FINAL 15/04/8

9 Policy implications and main messages Environmental improvements and innovation are key objectives of the Europe 2020 Strategy. 67 To reach these objectives it is essential to consider, monitor and assess the environmental impacts of different areas of EU policies throughout the entire policy cycle. Of course, almost all areas of policy have innovation consequences, since policy drives change in economic systems, and change will create new opportunities and incentives for innovators. In turn, such change will have environmental consequences. We focus on those areas of policy that are explicitly at the nexus of innovation and environment and are therefore of direct relevance to the work within EMInInn: (i) environmental policies; (ii) innovation policies; (iii) eco-innovation policies. •

•

•

While they are specifically aimed at the environment, environmental and energy policies often involve, unintentionally or deliberately, innovation effects. These are either a necessary condition to achieve the environmental objectives, or a technically necessary consequence having indirect implications for both the environment and the economy. Innovation policies (including R&D policies) are a key area of attention because they are generally not aimed at improving the environment and are dominated by growth and competitiveness objectives. They represent de facto a large part of what is defined as ‘industrial policy’ - but they are very likely to have an impact on the environment . This impact can be even a negative one, unless environmental concerns are embodied explicitly in innovation policy design and implementation. Eco-innovation policies, as emerged in the EU policy agenda during the last few years, explicitly recognise the direct or indirect (intentional/unintentional) links emerging from the above two policy areas and explicitly aims at directing innovation towards win-win outcomes of growth/competitiveness and environmental benefits.

The environmental implications of innovation can emerge and be very relevant for other policy areas, for example biodiversity and land-use, but the three policy areas we have identified display this relationship in a very direct and critical way. These three policy areas (and related areas, e.g. industrial policy) need specific tools to address the environmental implications of innovation, and such tools are required along the whole policy cycle: (i) design, including ex ante impact assessment (“tools of policy formulation”(Jordan and Turnpenny, 2015)); (ii) implementation, including monitoring; (iii) ex post evaluation. The results of EMinInn can offer support to these policy needs in three main areas: (i) (ii) (iii)

how to improve the understanding of how innovation and environmental performance are related, and how this relationship can be measured and monitored; how to include the environmental impact of innovation in ex ante policy design, in impact assessment and in ex post policy evaluation; how to improve ante macro-modelling for policy support through a better account of innovation and its environmental pressures.

67

Targets in Europe 2020: Reducing greenhouse gas emissions 20% (or even 30%, if the conditions are right) lower than 1990, 20% of energy from renewables, and 20% increase in energy efficiency

114

FINAL 15/04/8 The following sections provide a summary of the contributions of EMInInn to these policy tasks.

9.1 Understanding and measuring the relationship between innovation and environmental performance Developing indicators that describe the macro-environmental consequences of innovation requires an understanding of the mechanisms by which environmental outcomes, innovation, and policies interact. Classic economic approaches note that markets alone will not provide adequate incentives for eco-innovation, because of the “double externality” (related to the fact that both environmental pollution and innovation spill-overs are externalities). Hence, policy has a key role to play. But in practice, the complex interplay of various policy and market drivers and barriers of innovation in general, and eco-innovation in particular, creates challenges in identifying stable indicators that link the micro-level of innovation processes and macro-level of environmental outcomes. It is these challenges that EMInInn has sought to address. In doing so, EMInInn has provided a rich contribution to the understanding of these relationships – identifying the way in which policy induces both invention and innovation diffusion, for example. Various analytical results of EMInInn are about the use of indicators, both as such and within different types of models, to highlight the environmental implications of innovation, in particular innovation diffusion. The analysis has indicated that micro-level assessments do not always provide a full picture of the anticipated outcomes of the diffusion of an innovation. The wider conditions matter. This has implications for attempts to monitor eco-innovation: methods that rely on input and direct output indicators may measure the capacity or potential for eco-innovation—but may not always provide evidence that the innovation process is indeed resulting in environmental improvements. Key lessons are as follows: -

-

The relationship between policies, innovation processes and environmental outcomes is complex, and cannot easily be reduced to a single indicator. Eco-innovation is a relative concept: the same innovation would be considered an ecoinnovation in one context, but not in another. An example might be solar PV: it would certainly be an eco-innovation in Australia (a country with high insolation, and largely fossil-fuel based electricity), but perhaps not in Iceland (a country with low insolation and 100% renewable electricity). This relative nature of eco-innovation has important implications for how we understand monitoring activities, and in particular comparisons across space and time. The ‘greenness’ of an eco-innovation is not fixed. Care should thus be taken in constructing indices of eco-innovation that are based on technology-specific metrics (such as diffusion of particular technologies) and used to compare ‘eco-innovation’ across countries in which the actual environmental impacts of diffusion may vary considerably. The environmental outcomes associated with the diffusion of innovations may depend on the wider framework conditions—particularly environmental regulations and taxes—since unpredictable economic feedbacks may undermine the direct savings associated with the 115

FINAL 15/04/8 diffusion of an innovation that is environmentally beneficial at the micro-level. Incorporating rebound assessments into monitoring may be challenging—but one could envisage a system of indicators that incorporates measures of environmental policy stringency. Environmental policies can be expected to both induce green innovation and ensure that the diffusion of resulting innovations does actually result in environmental savings.

9.2 Improving environmental assessments in policy design and ex post policy evaluation The major challenge of effective technology-specific innovation policy (such as support for the deployment of specific technologies) is often thought to be uncertainty about whether a particular technology will become cost effective and profitable without the need for subsidy. However, determining the environmental pressures that might arise from diffusion of a technology has also been revealed to be very challenging. Too often, the environmental desirability of a particular technology is asserted rather than assessed. Of particular relevance is that the ultimate environmental effects of many technologies depend on the policy and market environment in which they are diffused. The environmental consequences of a technology are not solely an attribute of the technology itself, and a broader assessment is required that highlights how policymakers can ensure that environmental policy goals are met. Inter-sectoral influences are often forgotten or ignored in impact assessments. Consequently it is unclear whether the time and budgets are used in the most effective way and policy makers are regularly negatively surprised by unwanted and unexpected side effects. A policy might also be negatively assessed when no effect is visible within the sector but it turns out to generate the right outcome in another sector (for instance taxation on a price-inelastic activity which reduces the available household budget for other expenditures). To include the expected innovation effects of a policy and then the expected environmental implications of innovation can be important for good policy design—reducing the prevalence of unintended effects. Impact assessments associated with EU policy proposals are generally focused on the issue at stake and the expected economic consequences of the proposed policy, with limited consideration of indirect effects on the environment as mediated by innovation (be it at work outside the policy scope or induced by the policy itself, intentionally or not). Even where policies focus on inducing innovation (or even deploying specific technologies) the treatment of environmental consequences of that policy often neglects key processes, particularly rebound effects. The analysis within EMInInn suggests that improvements to such processes are possible, and offers a range of analytic tools for assessing policies. EMInInn has also produced a conceptual framework that helps policy makers to think about how innovation affects the environment (shown in Figure 9), and this can help to identify the processes that may need to be included within assessment processes. In particular, the EMInInn analysis leads to the following recommendations for Impact Assessments, monitoring and ex-post evaluation. However, it is clear that such detailed and therefore costly assessments should only be carried out in proportion to the expected outcome of the policy:

116

FINAL 15/04/8 •

•

• •

•

•

•

•

Authors of IAs and ex-post evaluations should be clear about limitations in their approach. Where uncertainties exist, information should be provided on how they might affect the outcome of the assessment, for example by conducting sensitivity analysis. A positive example where this has been done is the ex-post evaluation of the CIP Eco-Innovation market replication programme. Triangulation can be a useful tool to confirm the reliability of information. With triangulation a range of different methods, such as modelling, expert assessments, public consultation, surveys or case studies, are used in parallel. While modelling is used widely in many studies these days, it can provide a false sense of security. It is therefore advisable to counter-check model assumptions and results against other sources, such as expert opinions or public consultation. The case of the EU’s biofuels policy shows that failing to interrogate model assumptions can contribute to a failure to anticipate unintended consequences. When assessing the impact of innovation and technologies effects across the entire life-cycle of products should be considered. Also potential trade-offs between environmental pressures should be considered, as positive effects in one environmental area might lead to negative effects in another area. Mostly policy makers focus on reducing environmental pressures in a given priority area that the policy proposal aims to address and therefore tend to oversee potential negative environmental effects in other areas. This was for example the case with the EU’s biofuels policy. EMInInn has shown that while rebound effects can significantly offset environmental benefits on the marco-level, they are generally not considered in IAs and ex-post evaluations. Policy makers should therefore make use of the tools developed by EMInInn to take account of rebound effects in their assessment of policy options. If EU policy is being implemented at Member State level it is important to streamline monitoring activities in countries to ensure data comparability and therewith provide the basis for sound ex-post evaluation. The recent history of plant-by-plant data collection under the Large Combustion Plant Directive emphasizes this point: data submissions in early years used different plant naming conventions to later years, creating significant obstacles for attempts to use the data for monitoring the contribution of specific technologies to emissions abatement. When reporting environmental savings associated with specific technologies (or technologyspecific policies), a screening of potential rebound effects should be undertaken in order to avoid over-stating the expected contribution of the technologies. Where this takes place, it would be desirable to report savings under a range of future environmental pricing scenarios. For example, expected emissions savings associated with measures to promote energy efficiency might be reported under a range of carbon price scenarios, taking into account the way in which environmental consequences of rebound effects differ under different pricing regimes. Providing guidance for policy makers and external consultants that are responsible for assessing eco-innovation is key to improve current policy practice. The Commission should therefore actively compile, communicate and promote available guidance material. For example the “Guidebook to assessing environmental impacts of research and innovation policy” could be uploaded on the Commission’s website and promoted more actively in the policy evaluation community.

9.3 Conclusions Innovation policy is clearly an important strategy for achieving environmental policy goals costeffectively. However, eco-innovation does not necessarily provide an ‘easy way out’. Green innovations that save money are looked upon with great favour by users and policy makers alike 117

FINAL 15/04/8 because they constitute a win-win option. But it is not always straightforward to identify those innovations, and in any case the environmental wins may be low because of rebound effects. Currently used analytic tools that assess the environmental impacts of specific technologies rarely account for the full range of potential indirect effects. Using improved technology assessment tools, including those developed within EMInInn, may change the assessment of the expected environmental savings associated with specific technological development plans, or policies that are expected to result in the diffusion of particular technologies. In particular, analytic tools that incorporate economic feedback effects and potential life-cycle effects can result in the identification of issues that are sometimes neglected within currently dominant approaches to technology assessment in technology policy formulation. A key point is that the environmental performance of a given technology is not solely an inherent feature of the technology, but depends on wider conditions. In particular, the environmental consequences of ‘rebound effects’ will depend on carbon and energy prices, and policy options to address rebound effects need to be carefully considered. Current policy practice in impact assessment rarely makes such dependencies explicit. It makes sense for technology assessments to make clear where there are such dependencies, and to report a range of potential benefits of an eco-innovation policy package depending on the wider policy context, for example under differing carbon price scenarios. Most fundamentally, however, the analysis makes clear that while ecoinnovation policy is an important part of environmental policy, it needs to complemented by other environmental policy instruments to ensure that it yields its potential environmental benefits. Indicators of innovation system activity, such as patenting, may only provide a partial picture of the trends and drivers of the ultimate contribution of eco-innovation to reductions in environmental pressures. This is partly because such measures rely on an ability to distinguish ‘green’ innovation activities, and partly because of the complexity of the relationship between innovation system activities and environmental outcomes.

118

FINAL 15/04/8

10 Annex on results of macro-modelling: rebound effect and sensitivity testing 10.1 Rebound effects 10.1.1 Introduction

In this chapter we analyze the direct and indirect rebound effect(s) of eco-innovations68. Extensive research has yielded reasonable empirical basis to support the existence and relevance of the so-called rebound effect (Brookes (1990), Khazzoom (1980), Saunders (1992)). Comprehensive summaries of the current evidence can be found in (Sorrell, 2007) and Jenkins et al. (2011). In short, the rebound effect can be defined as the environmental savings that are thwarted or enhanced as a consequence of behavioral and systemic responses to technical changes. Because technology-based or engineering estimates of environmental savings generally fail to capture such responses, the actual savings might differ once these are incorporated. The rebound effect is thus commonly understood as the difference between the engineering and the actual savings, that is, the savings that are “taken back”. Barker et al. (Barker et al., 2007) states that “the idea that some or all of the expected reductions in energy consumption as a result of energy efficiency improvements are offset by increasing demand for energy services, arising from reductions in the effective price of energy services resulting from such improvements.” Therefore we measure the direct rebound effect by comparing the initial decrease of energy consumption caused by this eco-innovation to the final energy consumption found after the simulation. The indirect rebound effects are measures by the change in consumption goods and services due to this energy efficient eco-innovation (where both prices and income are kept fixed). We assume that the money saved will be redistributed among the other consumption goods and services. For this we compare two approaches; firstly, a partial equilibrium approach that provides us with both direct rebound effects, and secondly, the Computable General Equilibrium approach used in Deliverable 9.2, that includes both indirect and direct rebound effects. The exogenous shock we introduce in this chapter is based on the direct effects of the technical improvements/eco-innovations used in the computable general equilibrium model EXIOMOD. 10.1.2 Direct and indirect rebound effects

We want to measure the effect on the economy of the exogenous shock � Ci ,c ,inc for several consumption goods and income quintiles. The size of this exogenous shock is equal to the direct effects caused by the technical improvements/eco-innovations presented in D9.2. 69 We use a partial equilibrium approach to measure the direct and indirect rebound effect on consumption. We define the direct rebound effect as the different between the measure

Eco-innovations are innovations whose use is associated with less negative environmental impacts than conventional technologies (i.e. goods), see Arundel and Kemp (2009). In this project the environmental impact of the innovation is determined on a lifecycle basis. 69 Hence, this is the same direct effect as used in the setup of the full computable general equilibrium model EXIOMOD presented in D9.2 68

119

FINAL 15/04/8

shocks in the simulation and the exogenous shock. Moreover, we measure static InputOutput indirect effect of this consumption change on sales. The shock (consumption or demand change) is defined as:

� Ci ,c = ∑ � Ci ,c ,inc

(1)

inc

and the relative change is given by

%� Ci ,c =

� Ci ,c , � i ,c C

(2)

� i ,c denotes the level of consumption in country i of good or service c in the Where C reference scenario. The rebound effect is than defined as: = Ri ,c � Cisim ,c −� Ci ,c ,

(3)

where � Cisim ,c denotes the simulation results of the change in consumption in country i of good or service c . For all the consumption goods and services for which � Ci ,c ≠ 0 we call the above effect the direct rebound effect and for all other goods and services this is called the indirect rebound effect. 10.1.3 The partial equilibrium approach: indirect and direct rebounds

The partial equilibrium approach considers the changes in household demand that are caused by implementation of various technical improvement options under the assumption that neither prices nor incomes of the households are affected. In the partial equilibrium model prices and incomes are kept exogenously fixed and the households’ demand for various goods and services is derived based on the system of demand equations that give the solution to the utility maximization problem. In the current version of EXIOMOD a household in each income category maximizes the following Cobb-Douglas utility function (where α iH,* denotes the households power parameter):

U= i ,inc

∏ Ci,ci ,,cinc ⋅ Ti,inci ,T ⋅ H i,inci ,H , αH

αH

αH

(4)

c

where,

Ci ,c ,inc : consumption of goods/services c in country/region i for income group inc , H i ,inc , consumption of household goods/services in country/region i for income group inc , and

Ti ,inc , consumption of transport goods/services in country/region i for income group inc . Subject to its budget constraint:

120

FINAL 15/04/8

CBUDi ,inc= Yi ,inc ⋅ (1 − tyi ) + TRFi ,inc + HTRROWi ,inc ⋅ ERi − SH i ,inc ,

(5)

with

Yi ,inc the household income in country/region i for income group inc ,

tyi the income tax-rate, TRFi ,inc the social benefits from the government,

HTRROWi ,inc the transfers from the rest of the world

ERi the exchange rate and

SH i ,inc household savings per income group.

Representation of households’ behavior via this type utility function allows to reflect the differences in consumption structures of the households with different income levels. Maximization of the utility problem (4) yields the following optimal consumption levels: C= α iH,c ,inc ⋅ i ,c ,inc

= Ti ,inc α iH,T ,inc ⋅

CBUDi ,inc ( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c )

CBUDi ,inc

= H i ,inc α iH, H ,inc ⋅

PTi

,

CBUDi ,inc PH i

,

(6)

(7) ,

(8)

where trmi ,c denotes the trade and transport margins, PTM i the price of trade and transport margins (mathematical formulation follows later) and tci ,c the commodity tax. Due to inclusion of representation of three education levels and households grouped into five income quintile classes one can trace the effects of a specific policy on income redistribution and distinguish between measures that lead to increased and decreased income inequality in each specific country. To measure the rebound effect we introduce the (consumption or demand) shock into (6), (7) and (8) as follows:

Ci ,c ,inc

 CBUDi ,inc − ∑ � Ci ,c ,inc ⋅ ( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c )    +� C c , = α ⋅ i ,c ,inc   ( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c )     H i ,c ,inc

121

(9)

FINAL 15/04/8

Ti ,inc

CBUDi ,inc − ∑ � Ci ,c ,inc ⋅ ( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c ) c α = ⋅ +� Ci ,T ,inc , PTi

H i ,inc

H i ,T ,inc

CBUDi ,inc − ∑ � Ci ,c ,inc ⋅ ( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c ) c = α ⋅ +� Ci , H ,inc . PH i H i , H ,inc

(10)

(11)

The general pricing formula for the consumption goods is given by

(

σ Ai ,c σ Ai ,c 1 Pi ,c = ⋅ ( γ A1i ,c ⋅ PM i ,c ) + ( γ A2i ,c ⋅ PXDDEi ,c ) aAi ,c

)

1 1−σ Ai ,c

(12)

This can be interpreted as a weighted sum between prices of goods produced domestically ( PXDDEi ,c ) and prices of imported goods ( PM i ,c ). The scale parameter of Constant Elasticity of Substitution (CES) function of Armington function is denoted by aAi ,c and the Armington elasticity of substitution by σ Ai ,c . The formulation of trade flows of the EXIOMOD model is based on the Armington assumption of heterogeneity between the goods and services produced by different sectors and by different countries. According to this assumptions, similar commodities produced by different producers represent unique varieties which fall into the same category of goods or services specified in the model, and there is a demand for each of these unique varieties. The Armington specification is used in order to explain why one country would export and import at the same type commodities from the same category. One can imagine that all the trade of each commodity type in the country is coordinated by one trading agent, who doesn’t bear any additional cost and doesn’t require any commission. The trading agent operates in three steps. Firstly, it combines the commodities produced by different sector into one composite domestically produced product. Secondly, he determines the shares of this composite product which will be sold domestically and which will go on export. And lastly, he combines the domestic product with imported varieties and delivers the final product for intermediate consumption, consumption of government and households and for fixed capital formation.

122

FINAL 15/04/8

Figure 24 Structure of consumption nest

Transport output nest In the current version of EXIOMOD consumers can choose between different types of transport. On the top level the choice between own and purchased transport goods and services. Own transportation consists of fuel use, capital use (car capital stock), maintenance costs etc. There is a substitution possibilities between fuel use and capital stock of these sector. For simplification reasons we drop the income index inc . Level I of the transport output nest The demand function for own ( TOi ) and purchased transport ( TPi ) is modelled using a CES function in the following way σ Ti

 γ Oi  TOi = Ti ⋅    PTOi 

⋅ PTiσ Ti ⋅ aTiσ Ti −1 ,

(13)

σ Ti

 γP  TPi = Ti ⋅  i   PTPi 

⋅ PT σ Ti ⋅ aTiσ Ti −1 ,

(14)

with, 1

1−σ Ti 1  , PTi =⋅  ∑ γ jσ Ti ⋅ Pi , j1−σ Ti  aTi  j 

(15)

123

FINAL 15/04/8

for j = {O, P} , Pi ,O = PTOi and Pi , P = PTPi . Where γ Oi and γ Pi denote share parameters for the own and purchased transport, σ Ti the elasticity of substitution in the transport nest and aTi the scaling parameter in the transport nest.

Finally the price of own transport and purchased transport is given by = PTOi +

∑ ioc

⋅ Ti ⋅ ( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c ))

∑ ioc

c∈CET

PTPi =

i ,c

c∈CTO

i ,c ⋅ ETi ⋅ ( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c ))

∑ ioc

c∈cTP

i ,c

⋅ TPi ⋅ ( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c ))

,

(16)

,

(17)

where ioci ,c stands for the (Leontief) technical coefficients of intermediate inputs , CET , CTO and CTP the (sub)set of commodities and services bellowing to energy types used for transport, own transport and purchased transport (public transport), respectively. Level II of the transport output nest

KETi = θiKET TOi ,

(18)

(

1 PKETi = ⋅ γ ETiσ KETi ⋅ PETi1−σ KETi + γ KTiσ KETi ⋅ PKTi1−σ KETi aKETi

)

1 1−σ KETi

,

(19)

where θiKET determines the share of Own transport that comes from KETi , γ ETi and γ KTi denote share parameters for capital transport and energy transport, σ KETi the elasticity of substitution in the capital-energy-transport nest and aKETi the scaling parameter in the capital-energy-transport nest. Level III of the transport output nest σ KETi

 γ KTi  KTi = KETi ⋅    PKTi 

⋅ PKETiσ KETi ⋅ aKETiσ KETi −1 ,

PKTi =PKTi 0 =ri0 + δ i0,T ⋅ PI i0 ,

(20) (21)

σ KETi

 γ KTi  ETi = KETi ⋅    PETi  = PETi

∑ ioc

c∈CET

i ,c

⋅ PKETiσ KETi ⋅ aKETiσ KETi −1 ,

⋅ ETi ⋅ ( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c )) .

Housing output nest 124

(22) (23)

FINAL 15/04/8

Level I housing output nest

∑ ioc

PH i =

c∈CH

+

i ,c

⋅ ( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c ))

∑ ioc

c∈CEH

i ,c

⋅ ( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c ))

,

(24)

Level II housing output nest

KEH i = θiKEH H i

(25)

(

1 PKEH i = ⋅ γ EH iσ KEH i ⋅ PEH i1−σ KETi + γ KH iσ KETi ⋅ PKH i1−σ KETi aKEH i

)

1 1−σ KEH i

(26)

where θiKEH determines the share of housing that comes from KETi , γ EH i and γ KH i denote share parameters for capital and energy in housing, σ KEH i the elasticity of substitution in the capital-energy-housing nest and aKEH i the scaling parameter in the capital-energy-housing nest.

125

FINAL 15/04/8

Level III housing output nest σ KEH i

 γ KH i  KH i = KEH i ⋅    PKH i 

⋅ PKEH iσ KEHi ⋅ aKEH iσ KEHi −1 ,

PKH i =PKH i0 =ri0 + δ i0, H ⋅ PI i0 ,

With ri0 the (initial) return on capital and

(27) (28)

δ i0,H the (initial) capital depreciation rate.

σ KEH i

 γ KH i  EH i = KEH i ⋅    PEH i  = PEH i

∑ ioc

c∈CEH

i ,c

⋅ PKEH iσ KEHi ⋅ aKEH iσ KEHi −1 ,

⋅( Pi ,c + trmi ,c ⋅ PTM i ) ⋅ (1 + tci ,c )) ,

(29) (30)

Where CEH denotes the (sub)set of commodities that below to energy housing. 10.1.4 The Input-Output formulation (indirect effects on sales)

For the analysis of the indirect rebound effect of policy scenarios often Input-Output models are used. These models describe national economies and describe the interactions that take place between the various sectors of the economy, such as energy, manufacturing and agriculture. These sectors produce goods and services and which are sold/consumed/used by other sectors (intermediate consumption) and to final demand categories, such as household consumption, investment, government consumption and exports. Input-Output models are based on Input-Output tables, these tables shows how many inputs each sector is using and how much output is produced.

� i ,c ,inc , Initial value of final demand is given by the sum of initial household consumption C � i ,c , initial investments I i ,c , initial changes in initial government consumption CG � i ,c and the initial consumption of services for production of transport and inventories SV � i ,c , i.e: trade margins TMX �= FD i ,c

� = ∑ FD ∑ C� i ,c ,inc

inc

i ,c ,inc

� i ,c + I i ,c + SV � i ,c + TMX � i ,c + CG

(31)

inc

We introduce the (consumption of demand) shock in the equation for total sales: X= i ,c

� i ,c , s IO � i ,c +� C ∑s �XDc,s ⋅ XDi,s + FD i ,c

(32)

Hence, total sales is equal the initial share of the Leontief Input-Output coefficients relative to initial domestic output multiplied by domestic output summed over all sectors plus initial final demand and the (consumption or demand) shock. The equation for total demand or domestic output is equal to total sales multiplied by the share of initial domestic production in initial total sales, multiplied by the share of domestic 126

FINAL 15/04/8

production of commodity c by sector in total domestic production. Equation (32) and (33) form a system of simultaneous equations.

= XDi , s

∑ X i ,c , c

� XXD i ,c � XDD i ,c , s ⋅ � X i ,c ∑ � XDD i , ss

(33)

ss

127

FINAL 15/04/8 10.1.5 Simulation results for eco-innovations analyzed in the Built Environment WP

In this section we present simulation results for the Built Environment WP. As mentioned earlier, the shocks introduces in the simulation are based on the direct effects of the technical improvements/eco-innovations used in the computable general equilibrium (CGE) model EXIOMOD. Figure 24 shows a clear rebound effect, i.e. there exists a consumption shift compared to the baseline. The eco-innovations have opposite effects on the different product groups. For example, due to the eco-innovations less ‘coal and lignite, crude oil and gas’ is used but more ‘food, beverages and tobacco’. This could mean that the money the consumer saved from using less ‘energy’ is spend on other products. Moreover, the ‘final’ direct effect caused by eco-innovations measures using a partial equilibrium model are lower than the initial shock. This hold for all three categories.

2,00%

Agricultural products Products of forestry Coal and lignite, crude oil and gas Metal ores Non-metalic minerals Food,beverages and tobacco Textiles, leather, wood and paper products Coke, gas and coal Motor fuels and other petroleum products and… Plastic, rubber, glass, ceramics and chemicals Construction products Metal products Machnery, equipment, electronics and furniture Electricity Sale, trade, distribution and hotel Transportation Other private services Other public services Waste incineration and biogasification Waste landfill and composting

Consumption change (2050)

2050 BuiltEnv_PE 2050 BuiltEnv_shock

0,00% -2,00% -4,00% -6,00% -8,00% -10,00% -12,00% -14,00% -16,00%

Figure 25 Change in consumption due to the eco-innovations analysed in the Built Environment WP relative to the baseline.

128

FINAL 15/04/8

In Table 15 we present results for the consumption change for the years 2020 and 2050.

2020

2020

2050

2050

shock

PE

shock

PE

Agricultural products

0.1002 %

0.3004%

Products of forestry

0.0890 %

0.2629%

4.3325 %

Coal and lignite, crude oil and gas

4.2405 %

13.7262 %

13.4470 %

Metal ores

0.0692 %

0.1685%

Non-metalic minerals

0.0917 %

0.2785%

Food,beverages and tobacco

0.1020 %

0.3081%

Textiles, leather, wood and paper products

0.1038 %

0.3143%

Coke, gas and coal

0.0907 %

0.2684%

Motor fuels and other petroleum products and nuclear fuel

0.1015 %

0.3105%

Plastic, rubber, glass, ceramics and chemicals

0.1025 %

0.3123%

Construction products

0.1024 %

0.3107%

Metal products

0.1004 %

0.3079%

Machnery, equipment, electronics and furniture

0.0959 %

0.2952%

3.1311

Electricity

129

3.0261

-8.7626%

-8.4360%

FINAL 15/04/8

Sale, trade, distribution and hotel

%

%

0.3775 %

0.2821 %

-1.2478%

-0.9599%

Transportation

0.0964 %

0.2955%

Other private services

0.0999 %

0.3077%

Other public services

0.0927 %

0.2866%

Waste incineration and biogasification

0.1079 %

0.3356%

Waste landfill and composting

0.1094 %

0.3350%

Table 15 Change in consumption relative to the baseline caused by the eco-innovations in the Transport WP

Figure 25 shows the change in material use compared to the baseline. The simulation results for the CGE and PE. We see that the change in fuels (all the energy sources except electricity) and metals for the CGE is larger than resulting from the PE simulations. Since a CGE model captures wider economic effects, i.e. both direct and indirect effect of the ecoinnovations and the PE model only measures the direct effect. However, we must state that the CGE includes the changes in production (in each sector) and the PE only includes changes in production plus the changes in intermediate consumption between sectors caused by the input output linkages. The same holds for minerals, however the difference between the CGE and the PE results is smaller. For Biomass the opposite holds. It seems that some of the direct savings are lost when we look at the economic wide effects, i.e. also the indirect effects. In Table 16 an overview for material use is given for the years 2020, 2030, 2040 and 2050.

130

FINAL 15/04/8

Change in material use Fuels

Biomass

Metals

Minerals

0,40% 0,20% 0,00% -0,20%

2050 BuiltEnv

-0,40%

2050 BuiltEnv_PE

-0,60% -0,80% -1,00% -1,20% -1,40%

Figure 26 Change in material use relative to the baseline for 2050 both CGE and PE approach

Fuels

Biomass

Metals

Minerals

2020

BuiltEnv

-0.4269%

0.0204%

-0.1357%

0.0078%

2020

BuiltEnv_PE

-0.3377%

0.0537%

-0.1183%

-0.0139%

2030

BuiltEnv

-0.6900%

0.0572%

-0.2907%

-0.0059%

2030

BuiltEnv_PE

-0.4176%

0.0879%

-0.1854%

-0.0211%

2040

BuiltEnv

-0.9342%

0.0908%

-0.3797%

-0.0177%

2040

BuiltEnv_PE

-0.4626%

0.1175%

-0.2435%

-0.0270%

2050

BuiltEnv

-1.1381%

0.1237%

-0.4404%

-0.0358%

2050

BuiltEnv_PE

-0.4767%

0.1407%

-0.2845%

-0.0314%

Table 16 Change in material use for 2020, 2030, 2040 and 2050 relative to the baseline for CGE and PE approach

Finally, we look at the change in production emissions. These are presented in Figure 29. When we compare the PE results with the CGE results, we see that some of the environmental gains (less emissions) resulting from the eco-innovations is lost, i.e. the PE show lower emissions relative to the baseline, than the CGE results.

131

FINAL 15/04/8

Changes in production emissions 2020

2020

2050

2050

BuiltEnv

BuiltEnv_PE

BuiltEnv

BuiltEnv_PE

0,00% -0,20% -0,40%

GHG

-0,60%

Non-GHG

-0,80% -1,00% -1,20% -1,40% Figure 27 Relative changes in production emissions for CGE and PE approach

132

FINAL 15/04/8 10.1.6 Simulation Results for eco-innovations analyzed in the Transport WP In this section we present simulation results for the Transport WP. Figure 27 also shows a

clear rebound effect, i.e. there exists a consumption shift compared to the baseline. The eco-innovations have opposite effects on the different product groups. For example, due to the eco-innovations less ‘motor fuels and other petroleum products and nuclear fuels’ is used but more ‘food, beverages and tobacco’. This could mean that the money the consumer saved from using less ‘energy’ is spend on other products.

Rebound effect (consumption)

0,02% 0,01%

Transport_PE Transport_shock

0,00% -0,01% -0,02% -0,03% -0,04% -0,05% -0,06% Figure 28 Change in consumption due to the eco-innovations analysed in the Transport WP relative to the baseline for 2050.

133

FINAL 15/04/8

In Table 17 we present results for the consumption change for the years 2020 and 2050.

134

FINAL 15/04/8

2020

2020

2050

2050

shock

PE

shock

PE

Agricultural products

0.0006%

0.0007%

Products of forestry

0.0003%

0.0004%

Coal and lignite, crude oil and gas

0.0005%

0.0076%

Metal ores

0.0000%

0.0000%

Non-metalic minerals

0.0001%

0.0003%

Food,beverages and tobacco

0.0007%

0.0009%

Textiles, leather, wood and paper products

0.0008%

0.0010%

Coke, gas and coal

0.0000%

0.0002%

Motor fuels and other petroleum products and -0.0189% nuclear fuel

-0.0395%

-0.0260%

-0.0507%

Plastic, rubber, glass, ceramics and chemicals

0.0008%

0.0013%

0.0023%

0.0000%

Construction products

0.0008%

0.0010%

Metal products

0.0009%

0.0011%

Machnery, equipment, electronics and furniture

0.0009%

0.0011%

Electricity

0.0002%

0.0016%

0.0045%

0.0060%

Sale, trade, distribution and hotel

0.0052%

0.0056%

Transportation

0.0136%

0.0143%

Other private services

0.0008%

0.0010%

Other public services

0.0008%

0.0009%

Waste incineration and biogasification

0.0003%

0.0005%

Waste landfill and composting

0.0010%

0.0011%

Table 17 Change in consumption relative to the baseline caused by the eco-innovations in the Transport WP

Figure 28 shows the change in material use compared to the baseline for the simulation

results of the CGE and PE. We see that the change in fuels, biomass, metals, and minerals use for the CGE is larger than resulting from the PE simulations. The relative PE results are 135

FINAL 15/04/8

very small (around zero, i.e. almost equal to the baseline). However, we must state that the CGE includes the changes in production (in each sector) and the PE only includes changes in production plus the changes in intermediate consumption between sectors caused by the input output linkages. In Table 18 an overview for material use is given for the years 2020, 2030, 2040 and 2050.

Change in material use Fuels

Biomass

Metals

Minerals

00.000% 00.000% 00.000% 2050 Transport

00.000%

2050 Transport_PE

00.000% 00.000% 00.000% 00.000%

Figure 29 Change in material use relative to the baseline for 2050 both CGE and PE approach

Fuels

Biomass

Metals

Minerals

2020

Transport

0.0098%

0.0110%

0.0851%

0.0164%

2020

Transport_PE

-0.0111%

0.0003%

0.0000%

0.0000%

2030

Transport

0.0395%

0.0240%

0.1714%

0.0349%

2030

Transport_PE

-0.0074%

0.0004%

0.0000%

0.0001%

2040

Transport

0.0723%

0.0411%

0.2342%

0.0671%

2040

Transport_PE

-0.0060%

0.0004%

0.0001%

0.0001%

2050

Transport

0.0927%

0.0846%

0.2408%

0.0249%

2050

Transport_PE

-0.0048%

0.0005%

0.0001%

0.0001%

Table 18 Change in material use for 2020, 2030, 2040 and 2050 relative to the baseline for CGE and PE approach

136

FINAL 15/04/8

Changes in production emissions 0,14% 0,12% 0,10% 0,08% GHG

0,06%

Non-GHG

0,04% 0,02% 0,00% Transport

Transport_PE

Transport

Transport_PE

2020

2020

2050

2050

Figure 30 Relative changes in production emissions

10.2 Impacts on imports and exports of EU countries The impacts on considered improvement areas on the exports and imports of EU28 countries are small. The largest impacts are associated with the Transport and Energy improvement areas. Transport has positive impact on both exports and imports whereas Energy has negative impact on both exports and imports. Negative impact of trade from Energy improvement area is due to reduction in the production of the electricity sector (solar panels of households lead to this reduction). Transport leads to more imports because of the need to purchase specific materials for the production of electric cars. Increase in exports can be explained by the rebound effects and higher household consumption that is translated into higher import of goods. The overall innovation package leads to negative impacts on both exports and imports. It should be stressed that effects are quite small.

137

FINAL 15/04/8

Relative changes in EU28 exports 0,15% 0,10% 0,05%

BuiltEnv Transport

0,00%

Energy

-0,05%

Waste ICT

-0,10%

All

-0,15% -0,20% 2020

2030

2040

2050

Figure 31 Relative changes in exports of EU28 countries compared to the baseline

Relative changes in EU28 imports 0,15% 0,10% 0,05%

BuiltEnv Transport

0,00%

Energy

-0,05%

Waste ICT

-0,10%

All

-0,15% -0,20% 2020

2030

2040

2050

Figure 32 Relative changes in imports of EU28 countries compared to the baseline

The total changes in exports and imports are almost equal which is due to the trade balance constraint in the model. The changes in imports and exports by type of good and service differ a lot between type of goods and countries.

138

FINAL 15/04/8

10.3 Sensitivity analysis We have performed two additional simulations for Built Environment improvement area in order to check the sensitivity of results of EXIOMOD to changes in behavioral parameters. We have varies the elasticity of substitution between capital and labour by multiplying and dividing it by two. The variations in the results are very small which indicates that the model results are robust. The changes in production emissions are more affected than the changes in GDP.

Relative changes in EU28 GDP 0,000% -0,002% -0,004% -0,006% -0,008% -0,010% -0,012% -0,014%

2020

2030

2040

2050

BuiltEnv

-0,007%

-0,009%

-0,011%

-0,012%

BuiltEnv sigmaK Ldivided by 2

-0,007%

-0,010%

-0,011%

-0,012%

BuiltEnv sigmaKL multiplied by 2

-0,007%

-0,009%

-0,011%

-0,012%

Figure 33 Relative changes in EU28 GDP compared to the baseline

139

FINAL 15/04/8

Relative changes in production emissions 0,000% -0,100% -0,200% -0,300% -0,400% -0,500% -0,600% -0,700% -0,800% -0,900% -1,000% BuiltEnv

BuiltEnv sigmaKL divided by 2

BuiltEnv sigmaKL multiplied by 2

2050

2050

2050

GHG

-0,931%

-0,932%

-0,932%

Non-GHG

-0,115%

-0,123%

-0,111%

Figure 34 Relative changes in production emissions compared to the baseline

10.4 Conclusions In this chapter we analyze the direct and indirect rebound effects of eco-innovations. For this we compare two approaches; firstly, a partial equilibrium (PE) approach that provides us with direct rebound effects, and secondly the Computable General Equilibrium (CGE) approach used in Deliverable 9.2., with both indirect and direct rebound effects. The exogenous shock we introduce in this chapter is based on the direct effects of the technical improvements/eco-innovations used in the computable general equilibrium model EXIOMOD. We see a clear consumption shift compared to the baseline in the partial equilibrium approach, for both the eco-innovations introduced in the Transport WP and the Built Environment WP. This consumption shift, i.e. a redistribution of consumer expenses, can also be seen as a rebound effect (indirect). So, the eco-innovations have opposite effects on different product groups. Moreover, in case of the eco-innovation introduced in the Built Environment WP, we see that the change in fuels and metals (and minerals, however effect is smaller) use for the CGE is larger than resulting from the PE simulations. Since a CGE model captures wider economic effects, i.e. both direct and indirect effect of the eco-innovations and the PE model only measures the direct effect. However, we must state that the CGE includes the changes in production (in each sector) and the PE only includes changes in production plus the changes in intermediate consumption between sectors caused by the input output linkage. For the eco-innovations analyzed in the Transport WP we see that the change in fuels, biomass, 140

FINAL 15/04/8

metals, and minerals use for the CGE is larger than resulting from the PE simulations. The relative PE results are very small (around zero, i.e. almost equal to the baseline). When we compare the PE results with the CGE results, for emissions, we see that some of the environmental gains (less emissions) resulting from the eco-innovations is lost, i.e. the PE show lower emissions relative to the baseline, than the CGE results. This can also be identified as a clear rebound effects. This also holds for the both the eco-innovations analyzed in the Transport WP and the Built Environment WP. Finally, we do a sensitivity analysis by varying the elasticity of substitution between capital and labour by multiplying and dividing it by two. The variations in the results are very small which indicates that the model results are robust. The changes in production emissions are more affected than the changes in GDP.

141

FINAL 15/04/8

11 Full list of EMInInn Deliverables and reports D1.1

Report on the input from the first stakeholder workshop

D1.2

Final report, also addressing the suitability of adapting selected indicators to monitor innovation

D1.3

Summary of the outcomes

D2.1

Report on the input from the first stakeholder workshop

D2.2

Report on drivers and barriers to eco-innovation

D2.3

Report linking energy- and material efficiency improvements and indicators to innovations

D3.1

Draft report on the survey of methods, tools and data relevant for the project

D3.2

Full draft report on the outcomes of tasks 2, 3 and 4

D3.3

Final report on the analytical framework filled in with methods and data

D4.1

Report on the diffusion of different energy technologies and its macro-environmental impacts

D4.2

Database of environmental impacts from energy technologies

D4.3

Macro-environmental assessment of different scenarios whereby an 80% carbon reduction is achieved

D5.1

Report on the diffusion of ICT and its macro-environmental impacts

D5.2

Result of the ex post and ex ante assessment of major new ICT

D6.1

Report on the diffusion of transport technologies and its macro-environmental impacts

D6.2

Result of the ex post and ex ante assessment of major new low carbon technologies

D7.1

Report on the diffusion of Innovations in the field of housing and its macro-environmental impacts

D7.2

Report of a macro-environmental assessment of different scenarios for carbon reduction

D8.1

Report on relevant innovations in waste management

D8.2

Report on the environmental impacts of waste-related innovations

D8.3

Report on policy drivers

D9.1

Report on diffusion scenarios 142

FINAL 15/04/8 D9.2

Report on direct and indirect environmental impacts

D9.3

Comparison of model results

D10.1

Report of the resource and waste policies

D10.2

Report on the climate and energy policies

D10.3

Report on the biodiversity and land use policies

D10.4

Report on innovation policies

D11.1

Project web page

D11.2

General promotion material for the project

D11.3

2nd workshop proceedings

D11.4

3rd workshop proceedings

D11.5

Outcomes of the thematic workshops

D11.6

Final conference proceedings

D11.7

Policy brief on resource and waste policies

D11.8

Policy brief on energy and climate change policies

D11.9

Policy brief on biodiversity and land-use policies

D11.10

Policy brief on innovation policies

Background Paper 1 Background Paper 2 Background Paper 3 Background Paper 4 Background Paper 5

Background Paper 1 for the EmInInn Ex-Post Framework of Analysis. EMInInn Glossary of Terms working document Background Paper 2 for the EmInInn Ex-Post Framework of Analysis. Defining the research questions, defining the system and interpreting the results Background Paper 3 for the EmInInn Ex-Post Framework of Analysis: Environmental Indicators for the EmInInn project Background Paper 4 for the EmInInn Ex-Post Framework of Analysis: Innovation indicators Background Paper 5 for the EmInInn Ex-Post Framework of Analysis: Methods and models for the EmInInn project

143

FINAL 15/04/8

12 References ABERNATHY, W. & CLARK, K. 1985. Innovation: mapping the winds of creative destruction. Research Policy, 14, 3-22. ACOST 1992. Cleaner Technology, London, HMSO. Advisory Council on Science and Technology. ARUNDEL, A., KEMP, R. & PARTO, S. 2007. Indicators for Environmental Innovation: What and How to Measure. In: MARINOVA, D., ANNANDALE, D. & PHILLIMORE, J. (eds.) International Handbook on Environment and Technology Management (ETM). Cheltenham: Edward Elgar. ASHFORD, N. 1993. Understanding Technological Responses of Industrial Firms to Environmental Problems: Implications for Government Policy. In: FISCHER, K. & SCHOT, J. (eds.) Environmental Strategies for Industry: International Perspectives on Research Needs and Policy Implication. Washington, DC: Island Press. BARKER, T., EKINS, P. & FOXON, T. 2007. The macro-economic rebound effect and the UK economy. Energy Policy, 35, 4935-4946. BARTOLOMEO, M., KEMP, R., RENNINGS, K. & ZWICK, T. 2003. Employment impacts of cleaner production: theory, methodology and results. In: RENNINGS, K. & ZWICK, T. (eds.) Employment Impacts of Cleaner Production. ZEW Economic Studies. Heildeberg, New York: Physica-Verlag, ZEW. BASTEIN, T., KOERS, W., DITTRICH, K., BECKER, J. & DIAZ LOPEZ, F. J. 2014. Business Barriers To The Uptake Of Eco-innovation and Resource Efficiency Measures. . Policy Options For A ResourceEfficient Economy, Deliverable D1.5. Delft: Netherlands Organisation for Applied Scientific Research TNO. BERGEK, A. & JACOBSSON, S. 2010. Are tradable green certificates a cost-efficient policy driving technical change or a rent-generating machine? Lessons from Sweden 2003-2008. Energy Policy, 38, 1255-1271. BERGEK, A., JACOBSSON, S., CARLSSON, B., LINDMARK, S. & RICKNE, A. 2008. Analyzing the functional dynamics of technological innovation systems: A scheme of analysis. Research Policy, 37, 407-429. BOWER, J. L. & CHRISTENSEN, C. M. 1995. Disruptive Technologies - Catching the Wave. Harvard Business Review, 73, 43-53. BRESNAHAN, T. F. & TRAJTENBERG, M. 1995. General purpose technologies ‘Engines of growth’? Journal of Econometrics, 65, 83-108. BUNZL, M. 2004. Counterfactual history: a user's guide. The American historical review, 109, 845858. BUSH, R., JACQUES, D. A., SCOTT, K. & BARRETT, J. 2014. The carbon payback of micro-generation: An integrated hybrid input–output approach. Applied Energy, 119, 85-98. COENEN, L., BENNEWORTH, P. & TRUFFER, B. 2012. Toward a spatial perspective on sustainability transitions. Research Policy, 41, 968-979. COENEN, L. & DÍAZ LÓPEZ, F. J. 2010. Comparing systems approaches to innovation and technological change for sustainable and competitive economies: an explorative study into conceptual commonalities, differences and complementarities. Journal of Cleaner Production, 18, 1149-1160. DEL RÍO, P., CARRILLO-HERMOSILLA, J. & KÖNNÖLÄ, T. 2010. Policy Strategies to Promote EcoInnovation. Journal of Industrial Ecology, 14, 541-557. DEL RIO, P., CARRILLO-HERMOSILLA, J., KÖNNÖLÄ, T. & BLEDA, M. In press. Business strategies and capacities for successful eco-innovation. Technological and Economic Development of Economy, 26. DIAZ LOPEZ, F. J. 2008. A tailored method for eco-innovation strategies and drivers. DIME International Conference “Innovation, Sustainability and Policy”. University Montesquieu Bordeaux IV, France.: DIME; GREThA, University Montesquieu Bordeaux IV.

144

FINAL 15/04/8 EC 2011. COMMUNICATION FROM THE COMMISSION TO THE EUROPEAN PARLIAMENT, THE COUNCIL, THE EUROPEAN ECONOMIC AND SOCIAL COMMITTEE AND THE COMMITTEE OF THE REGIONS. Innovation for a sustainable Future - The Eco-innovation Action Plan (Eco-AP). Brussels: Commission of the European Communitites. EDQUIST, C. 1997. Systems of Innovation: Technologies, Institutions and Organisations, London, Washington, Routledge. EDQUIST, C. 2005. Systems of Innovation. Perspectives and challenges. In: FAGERBERG, J., MOWERY, D. & NELSON, R. (eds.) The Oxford Handbook on Innovation. Oxford: Oxford University Press. EEA 2006. EN09 Emissions (CO2, SO2 and NOx) from public electricity and heat production – explanatory indicators. European Environment Agency, Copenhagen. . EKINS, P. 2010. Eco-innovation for environmental sustainability: concepts, progress and policies. International Economics and Economic Policy, 7, 267-290. EKINS, P., MCDOWALL, W. & ZENGHELIS, D. 2014. Greening the Recovery. The Report of the Green Economy Policy Commission. UCL. FONT VIVANCO, D., FREIRE-GONZÁLEZ, J., KEMP, R. & VAN DER VOET, E. 2014. The Remarkable Environmental Rebound Effect of Electric Cars: A Microeconomic Approach. Environmental Science & Technology, 48, 12063-12072. FREEMAN, C. 1995. The ‘National System of Innovation’in historical perspective. Cambridge Journal of Economics, 19, 5-24. FRONDEL, M., HORBACH, J. & RENNINGS, K. 2007. End-of-pipe or cleaner production? An empirical comparison of environmental innovation decisions across OECD countries. Business Strategy and the Environment, 16, 571-584. GALLAGHER, K. S., GRÜBLER, A., KUHL, L., NEMET, G. & WILSON, C. 2012. The Energy Technology Innovation System. Annual Review of Environment and Resources, 37, 137-162. GEELS, F. W. 2002. Technological transitions as evolutionary reconfiguration processes: a multi-level perspective and a case-study. Research Policy, 31, 1257-1274. HADJIMANOLIS, A. 2003. The Barriers Approacht to Innovation. In: L., S. (ed.) The International Handbook on Innovation. Elsevier Science Ltd. HEKKERT, M., SUURS, R., NEGRO, S., KUHLMANN, S. & SMITS, R. 2007. Functions of innovation systems: A new approach for analysing technological change. Technological Forecasting and Social Change, 74, 413-432. HORBACH, J., RAMMER, C. & RENNINGS, K. 2012. Determinants of eco-innovations by type of environmental impact: The role of regulatory push/pull, technology push and market pull. Ecological Economics, 78, 112-122. HUPPES, G., KLEIJN, R., HUELE, R., EKINS, P., SHAW, B., ESDERS, M. & SCHALTEGGER, S. 2008. ECODRIVE. A Framework for Measuring Eco-innovation: typology of indicators based on causal chains. EU FP6 funded project, Priority FP6-2005-SSP-5-A. Leiden: CML-Leiden University, PSI, CSM. JOHANSSON, G. & MAGNUSSON, T. 1998. Eco-innovations-a novel phenomenon? Journal of Sustainable Product Design, 7-18. JONES, E. & HARRISON, D. 2000. Investigating the use of TRIZ in Eco-innovation. JORDAN, A. & TURNPENNY, J. 2015. The tools of policy formulation: actors, capacities, venues and effects, Cheltenham, Edward Elgar. KEMP, R. 1997. Environmental Policy and Technical Change. A comparison of the Technological Impact of Policy Instruments., Cheltenham, Brookfield, Edward Elgar. KEMP, R. 2010. Eco-innovation: Definition, Measurement and Open Research Issues. Economia Politica, 397-420. KEMP, R. 2011. Ten themes for eco-innovation policies in Europe. S.A.P.I.EN.S, 4, 1-19. KEMP, R., DIAZ LOPEZ, F. J. & BLEISCHWITZ, R. 2013. Green Growth and Eco- innovation. Report of the FP7 project EMInInn Macro-level Indicators to monitor the environmental impact of

145

FINAL 15/04/8 innovation. Delft and Maastricht: Maastricht University, Netherlands Organisation for Applied Scientific Research TNO and Wuppertal Institute. KEMP, R. & FOXON, T. 2007. Typology of eco-innovation. Project Paper: Measuring Eco-Innovation, 1-24. KEMP, R. & PEARSON, P. 2007. Final report MEI project about measuring eco-innovation UM-MERIT. KEMP, R. & PEARSON, P. 2008. Final report MEI project about measuring eco-innovation. EU FP6 funded project 044513. Maastricht: UM-MERIT, ZEW, DTU, ICL, LEIA. KEMP, R. & SOETE, L. 1992. The greening of technological progress : An evolutionary perspective. Futures, 24, 437-457. KESIDOU, E. & DEMIREL, P. 2012. On the drivers of eco-innovations: Empirical evidence from the UK. Research Policy, 41, 862-870. KLEIN WOOLTHUIS, R., LANKHUIZEN, M. & GILSING, V. 2005. A system failure framework for innovation policy design. Technovation, 25, 609-619. KLEIN WOOLTHUIS, R. J. A. 2010. Sustainable Entrepreneurship in the Dutch Construction Industry. Sustainability, 2, 505-523. KLEMMER, P., LEHR, U. & LOBBE, K. 1999. Environmental Innovation. Incentives and barriers., Berlin, German Ministry of Research and Technology (BMBF). Analytica-Verlag. KOBAYASHI, Y., KOBAYASHI, H., HONGU, A. & SANEHIRA, K. 2005. A Practical Method for Quantifying Eco-efficiency Using Eco-design Support Tools. Journal of Industrial Ecology, 9, 131-144. KOBERG, C. S., UHLENBRUCK, N. & SARASON, Y. 1996. Facilitators of organizational innovation: The role of life-cycle stage. Journal of Business Venturing, 11, 133-149. LEHTONEN, M. 2015. Indicators: ‘tools’ for policy formulation or an inexhaustible source of unanticipated impacts? In: JORDAN, A. & TURNPENNY, J. (eds.) The Tools of Policy Formulation. . Cheltenham: Edward Elgar. MALERBA, F. 2004. Sectoral Systems of Innovation. Concepts, issues and analysis of six major sectors in Europe, Cambridge, Cambridge University Press. MARKARD, J., RAVEN, R. & TRUFFER, B. 2012. Sustainability transitions: An emerging field of research and its prospects. Research Policy, 41, 955-967. MAZZANTI, M. & ZOBOLI, R. 2009. Embedding environmental innovation in local production systems: SME strategies, networking and industrial relations: evidence on innovation drivers in industrial districts. International Review of Applied Economics, 23, 169-195. MICHAELIS, M. & CAREY, W. 1973. Barriers to Innovation in Industry: Opportunities for Public Policy Challenges. Washington D.C.: Arthur D. Little and Industrial Research Institute for the National Science Foundation. MONTALVO, C. 2008. General wisdom concerning the factors affecting the adoption of cleaner technologies: a survey 1990-2007. Journal of Cleaner Production, 16, S7-S13. MONTALVO, C., DIAZ LOPEZ, F. J. & BRANDES, F. 2011. Analysis of the Potential for Eco-innovation in Nine Sectors. Task 4 Horizontal Report 4. Delft: Europe Innova Sectoral Innovation Watch. Project on behalf of the European Commission, DG Enterprise and Industry. OECD 1985. Environmental Policy and Technological Change. Paris: : Organisation for Economic Cooperation and Development.OECD. OECD & EUROSTAT 2005. Guidelines for collecting and interpreting innovation data -Oslo Manual, Paris, Organisation for Economic Cooperation and Development. EUROSTAT. PAVITT, K. 1984. Sectoral patterns of technical change: towards a taxonomy and a theory. Research Policy, 13, 343-373. PIATIER, A. 1984. Barriers to Innovation. A study carried out for the Commission of the European Communities, London, Frances Pinter (Publishers) London and Dover NH. POPP, D. 2010. Innovation and Climate Policy. Annual Review of Resource Economics, 2, 275-298. RAJAGOPAL, D. 2014. Consequential Life Cycle Assessment of Policy Vulnerability to Price Effects. Journal of Industrial Ecology, 18, 164-175.

146

FINAL 15/04/8 RENNINGS, K. 2000a. Redefining innovation -- eco-innovation research and the contribution from ecological economics. Ecological Economics, 32, 319-332. RENNINGS, K. 2000b. Redefining innovation — eco-innovation research and the contribution from ecological economics. Ecological Economics, 32, 319-332. RENNINGS, K. 2014. Introduction: Global diffusion of environmental innovations. Environmental Innovation and Societal Transitions, 10, 1-3. RHODES, A., SKEA, J. & HANNON, M. 2014. The Global Surge in Energy Innovation. Energies, 7, 56015623. ROSENBERG, N. 1976. Technology and Environment: and economic exploration. Perspectives on Technology. London: Cambridge University Press. ROSENBERG, N. 1982. Inside the black box: technology and economics, Cambridge, Cambridge University Press. SKEA, J. 1995. Environmental technology. In: FOLMER, H. & GABEL, H. (eds.) Principles of Environmental and Resource Economics. A Guide for Students and Decision-Makers Second Edition. 2nd ed. Cheltenham, Northampton: Edward Elgar. SMIL, V. 2000. Perils of long-range energy forecasting: reflections on looking far ahead. Technological Forecasting and Social Change, 65, 251-264. SORRELL, S. 2007. The rebound effect: an assessment of the evidence for economy-wide energy savings from improved energy efficiency. Project Report. SUH, S. & HUPPES, G. 2005. Methods for life cycle inventory of a product. Journal of Cleaner Production, 13, 687-697. SUH, S., LENZEN, M., TRELOAR, G. J., HONDO, H., HORVATH, A., HUPPES, G., JOLLIET, O., KLANN, U., KREWITT, W. & MORIGUCHI, Y. 2004. System boundary selection in life-cycle inventories using hybrid approaches. Environmental Science & Technology, 38, 657-664. TAYLOR, M. R., RUBIN, E. S. & HOUNSHELL, D. A. 2005. Regulation as the mother of innovation: The case of SO 2 control. Law and Policy, 27, 348-378. TRIGUERO, A., MORENO-MONDÉJAR, L. & DAVIA, M. A. 2013. Drivers of different types of ecoinnovation in European SMEs. Ecological Economics, 92, 25-33. TUKKER, A., POLIAKOV, E., HEIJUNGS, R., HAWKINS, T., NEUWAHL, F., RUEDA-CANTUCHE, J. M., GILJUM, S., MOLL, S., OOSTERHAVEN, J. & BOUWMEESTER, M. 2009. Towards a global multiregional environmentally extended input-output database. Ecological Economics, 68, 19281937. VAN DEN BERGH, J. C. J. M. 2011. Energy Conservation More Effective With Rebound Policy. Environmental and Resource Economics, 48, 43-58. VAN DEN BERGH, J. C. J. M. 2013. Environmental and climate innovation: Limitations, policies and prices. Technological forecasting and social change, 80, 11-23. VAN DEN BERGH, J. C. J. M., TRUFFER, B. & KALLIS, G. 2011. Environmental innovation and societal transitions: Introduction and overview. Environmental Innovation and Societal Transitions, 1, 1-23. VIVANCO, D. F. & VAN DER VOET, E. 2014. The rebound effect through industrial ecology’s eyes: a review of LCA-based studies. The International Journal of Life Cycle Assessment, 19, 19331947. WEBER, L. 1997. Some reflections on barriers to the efficient use of energy. Energy Policy, 25, 833835. WETERINGS, R., KUIJPER, J., SMEETS E, ARNOKKE´E GJ & MINNE, B. 1997. 81 options, technology for sustainable development. Delft: TNO/STB.

147