Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Table of Contents 2 Chapter 16: Innovation, technology development and transfer ..................................................... 16-1 3 Executive summary........................................................................................................................ 16-4 4 16.1 Introduction ........................................................................................................................ 16-8 5 16.2 Elements, drivers and modelling of technology innovation ............................................. 16-11 6 16.2.1 Stages of the innovation process .............................................................................. 16-11 7 16.2.2 Sources of technological change .............................................................................. 16-14 8 Cross-Chapter Box 11: Digitalization: efficiency potentials and governance considerations . 16-18 9 16.2.3 Directing technological change ................................................................................ 16-22 10 16.2.4 Representation of the innovation process in modelled decarbonization pathways .. 16-25 11 16.3 A systemic view of technological innovation processes .................................................. 16-29 12 16.3.1 Frameworks for analysing technological innovation processes ............................... 16-29 13 16.3.2 Identifying systemic failures to innovation in climate-related technologies ............ 16-31 14 16.3.3 Indicators for technological innovation ................................................................... 16-35 15 16.3.4 Emerging policy perspectives on systemic transformations .................................... 16-42 16 16.4 Innovation policies and institutions ................................................................................. 16-45 17 16.4.1 Overview of policy instruments for climate technology innovation ........................ 16-45 18 16.4.2 The drivers and politics of national policies for climate change mitigation and adaptation 19 16-47 20 16.4.3 Indicators to assess the innovation, competitiveness and distributional outcomes of 21 policy instruments .................................................................................................................... 16-48 22 16.4.4 Assessment of innovation and other impacts of innovation policy instruments ...... 16-49 23 16.4.5 Trade instruments and their impact on innovation ................................................... 16-61 24 16.4.6 Intellectual property rights, legal framework and the impact on innovation ........... 16-62 25 16.4.7 Sub-national innovation policies and industrial clusters .......................................... 16-64 26 16.4.8 System-oriented policies and instruments ................................................................ 16-65 27 16.5 International technology transfer and cooperation for transformative change................. 16-65 28 16.5.1 International cooperation on technology development and transfer: needs and 29 opportunities ............................................................................................................................ 16-65 30 16.5.2 Objectives and roles of international technology transfer and cooperation efforts .. 16-67 31 16.5.3 International technology transfer and cooperation: recent institutional approaches 16-69 32 16.5.4 Emerging ideas for international technology transfer and cooperation ................... 16-73 33 16.6 Technological change and sustainable development ....................................................... 16-75 34 16.6.1 Linking sustainable development and technological change ................................... 16-75 35 Cross-Chapter Box 12 Transition Dynamics ........................................................................... 16-77 36 16.6.2 Sustainable development and technological innovation: Synergies, trade-offs and 37 governance ............................................................................................................................... 16-81 Do Not Cite, Quote or Distribute 16-2 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 16.6.3 Actions that maximise synergies and minimise trade-offs between innovation and 2 sustainable development .......................................................................................................... 16-84 3 16.6.4 Climate change, sustainability development and innovation ................................... 16-86 4 16.7 Knowledge gaps ............................................................................................................... 16-87 5 Frequently Asked Questions (FAQs) ........................................................................................... 16-89 6 References .................................................................................................................................... 16-91 7 8 Do Not Cite, Quote or Distribute 16-3 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Executive summary 2 Innovation in climate mitigation technologies has seen enormous activity and significant 3 progress in recent years. Innovation has also led to, and exacerbated, trade-offs in 4 relation to sustainable development. (high confidence). Innovation, can leverage action to 5 mitigate climate change by reinforcing other interventions. In conjunction with other enabling 6 conditions innovation can support system transitions to limit warming and help shift 7 development pathways. The currently widespread implementation of solar photovoltaic (solar 8 PV) and LEDs, for instance, could not have happened without technological innovation (high 9 confidence). Technological innovation can also bring about new and improved ways of 10 delivering services that are essential to human well-being. At the same time as delivering 11 benefits, innovation can result in trade-offs that undermine both progress on mitigation and 12 progress towards other sustainable development goals. Trade-offs include negative 13 externalities – for instance greater environmental pollution and social inequalities – rebound 14 effects leading to lower net emission reductions or even increases in emissions, and increased 15 dependency on foreign knowledge and providers (high confidence). Effective governance and 16 policy has the potential to avoid and minimise such misalignments (medium evidence, high 17 agreement). {16.1, 16.2, 16.3, 16.4, 16.5.1, 16.6} 18 A systemic view of innovation to direct and organize the processes has grown over the 19 last decade. This systemic view of innovation takes into account the role of actors, 20 institutions, and their interactions and can inform how innovation systems that vary 21 across technologies, sectors and countries, can be strengthened (high confidence). Where 22 a systemic view of innovation has been taken, it has enabled the development and 23 implementation of indicators that are better able to provide insights in innovation processes. 24 This, in turn, has enabled the analysis and strengthening of innovation systems. Traditional 25 quantitative innovation indicators mainly include R&D investments and patents. Systemic 26 indicators of innovation, however, go well beyond these approaches. They include structural 27 innovation system elements including actors and networks, as well as indicators for how 28 innovation systems function, such as access to finance, employment in relevant sectors, and 29 lobbying activities. For example, in Latin America, monitoring systemic innovation indicators 30 for the effectiveness of agroecological mitigation approaches has provided insights on the 31 appropriateness and social alignment of new technologies and practices. Climate-energy- 32 economy models, including integrated assessment models, generally employ a stylised and 33 necessarily incomplete view of innovation, and have yet to incorporate a systemic 34 representation of innovation systems {16.2, 16.2.4, 16.3, 16.3.4, 16.5, Table 16.7, Box 16.1, 35 Box 16.5}. 36 A systemic perspective on technological change can provide insights to policymakers 37 supporting their selection of effective innovation policy instruments (high confidence). A 38 combination of scaled-up innovation investments with demand-pull interventions can achieve 39 faster technology unit cost reductions and more rapid scale-up than either approach in isolation 40 (high confidence). These innovation policy instruments would nonetheless have to be tailored 41 to local development priorities, to the specific context of different countries, and to the 42 technology being supported. The timing of interventions and any trade-offs with sustainable 43 development also need to be addressed. Public R&D funding and support as well as innovation 44 procurement have shown to be valuable for fostering innovation in small to medium cleantech Do Not Cite, Quote or Distribute 16-4 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 firms. Innovation outcomes of policy instruments not necessarily aimed at innovation, such as 2 feed-in tariffs, auctions, emissions trading schemes, taxes and renewable portfolio standards, 3 vary from negligible to positive for climate change mitigation. Some specific designs of 4 environmental taxation can also result in negative distributional outcomes. Most of the 5 available literature and evidence on innovation systems come from industrialised countries and 6 larger developing countries. However, there is a growing body of evidence from developing 7 countries and small island developing states (SIDS) {16.4, 16.4.4.3, 16.4.4.4, 16.5, 16.7}. 8 Experience and analyses show that technological change is inhibited if technological 9 innovation system functions are not adequately fulfilled, this inhibition occurs more often 10 in developing countries. (high confidence). Examples of such functions are knowledge 11 development, resource mobilisation, and activities that shape the needs, requirements and 12 expectations of actors within the innovation system (guidance of the search). Capabilities play 13 a key role in these functions, the build-up of which can be enhanced by domestic measures, but 14 also by international cooperation (high confidence). For instance, innovation cooperation on 15 wind energy has contributed to the accelerated global spread of this technology. As another 16 example, the policy guidance by the Indian government, which also promoted development of 17 data, testing capabilities and knowledge within the private sector, has been a key determinant 18 of the success of an energy-efficiency programme for air conditioners and refrigerators in India. 19 {16.3, 16.5, 16.6, Cross-Chapter Box 12 in this chapter, Box 16.3} 20 Consistent with innovation system approaches, the sharing of knowledge and experiences 21 between developed and developing countries can contribute to addressing global climate 22 and sustainable development goals. The effectiveness of such international cooperation 23 arrangements, however, depends on the way they are developed and implemented (high 24 confidence). The effectiveness and sustainable development benefits of technology sharing 25 under market conditions appears to be determined primarily by the complexity of technologies, 26 local capabilities and the policy regime. This suggests that the development of planning and 27 innovation capabilities remains necessary, especially in least-developed countries and SIDSs. 28 International diffusion of low-emission technologies is also facilitated by knowledge spillovers 29 from regions engaged in clean R&D (medium confidence). 30 The evidence on the role of intellectual property rights (IPR) in innovation is mixed. Some 31 literature suggests that it is a barrier while and other sources suggests that it is an enabler 32 to the diffusion of climate-related technologies (medium confidence). There is agreement 33 that countries with well-developed institutional capacity may benefit from a strengthened IPR 34 regime, but that countries with limited capabilities might face greater barriers to innovation as 35 a consequence. This enhances the continued need for capacity building. Ideas to improve the 36 alignment of the global IPR regime and addressing climate change include specific 37 arrangements for least-developed countries, case-by-case decision-making and patent-pooling 38 institutions. {16.2.3.3, 16.5, Box 16.10} 39 Although some initiatives have mobilised investments in developing countries, gaps in 40 innovation cooperation remain, including in the Paris Agreement instruments. These 41 gaps could be filled by enhancing financial support for international technology 42 cooperation, by strengthening cooperative approaches, and by helping build suitable 43 capacity in developing countries across all technological innovation system functions 44 (high confidence). The implementation of current arrangements of international cooperation Do Not Cite, Quote or Distribute 16-5 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 for technology development and transfer, as well as capacity building, are insufficient to meet 2 climate objectives and contribute to sustainable development. For example, despite building a 3 large market for mitigation technologies in developing countries, the lack of a systemic 4 perspective in the implementation of the Clean Development Mechanism, operational since the 5 mid-2000s, has only led to some technology transfer, especially to larger developing countries, 6 but limited capacity building and minimal technology development (medium confidence). In 7 the current climate regime, a more systemic approach to innovation cooperation could be 8 introduced by linking technology institutions, such as the Technology Mechanism, and 9 financial actors, such as the financial mechanism. {16.5.3} 10 Countries are exposed to sustainable development challenges in parallel with the 11 challenges that relate to climate change. Addressing both sets of challenges 12 simultaneously presents multiple and recurrent obstacles that systemic approaches to 13 technological change could help resolve, provided they are well managed (high 14 confidence). Obstacles include both entrenched power relations dominated by vested interests 15 that control and benefit from existing technologies, and governance structures that continue to 16 reproduce unsustainable patterns of production and consumption (medium confidence). Studies 17 also highlight the potential of cultural factors to strongly influence the pace and direction of 18 technological change. Sustainable solutions require adoption and mainstreaming of locally 19 novel technologies that can meet local needs, and simultaneously address the Sustainable 20 Development Goals (SDGs). Acknowledging the systemic nature of technological innovation, 21 which involve many levels of actors, stages of innovation and scales, can lead to new 22 opportunities to shift development pathways towards sustainability. {16.4, 16.5, 16.6} 23 An area where sustainable development, climate change mitigation and technological 24 change interact is digitalisation. Digital technologies can promote large increases in 25 energy efficiency through coordination and an economic shift to services, but they can 26 also greatly increase energy demand because of the energy used in digital devices. System- 27 level rebound effects may also occur. (high confidence) Digital devices, including servers, 28 increase pressure on the environment due to the demand for rare metals and end-of-life 29 disposal. The absence of adequate governance in many countries can lead to harsh working 30 conditions and unregulated disposal of electronic waste. Digitalization also affects firms' 31 competitiveness, the demand for skills, and the distribution of, and access to, resources. The 32 existing digital divide, especially in developing countries, and the lack of appropriate 33 governance of the digital revolution can hamper the role that digitalization could play in 34 supporting the achievement of stringent mitigation targets. At present, the understanding of 35 both the direct and indirect impacts of digitalization on energy use, carbon emissions and 36 potential mitigation, is limited (medium confidence). {Cross-Chapter Box 11 in this chapter, 37 16.2} 38 Strategies for climate change mitigation can be most effective in accelerating 39 transformative change when actions taken to strengthen one set of enabling conditions 40 also reinforce and strengthen the effectiveness of other enabling conditions (medium 41 confidence). Applying transition or system dynamics to decisions can help policymakers take 42 advantage of such high-leverage intervention points, address the specific characteristics of 43 technological stages, and respond to societal dynamics. Inspiration can be drawn from the 44 global unit cost reductions of solar PV, which were accelerated by a combination of factors Do Not Cite, Quote or Distribute 16-6 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 interacting in a mutually reinforcing way across a limited group of countries (high confidence). 2 {Box 16.2, Cross-Chapter Box 10 in chapter 14} 3 Better and more comprehensive data on innovation indicators can provide timely insights 4 for policymakers and policy design locally, nationally and internationally, especially for 5 developing countries, where such insights are missing more often. Data needed include on 6 those that can show the strength of technological, sectoral and national innovation systems. It 7 is also necessary to validate current results and generate insights from theoretical frameworks 8 and empirical studies for developing countries contexts. Innovation studies on adaptation and 9 mitigation other than energy and ex-post assessments of the effectiveness of various 10 innovation-related policies and interventions, including R&D, would also provide benefits. 11 Furthermore, methodological developments to improve the ability of Integrated Assessment 12 Models (IAMs) to capture energy innovation system dynamics, and the relevant institutions and 13 policies (including design and implementation), would allow for more realistic assessment. 14 {16.2, 16.3, 16.7} 15 Do Not Cite, Quote or Distribute 16-7 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 16.1 Introduction 2 Technological change and innovation are considered key drivers of economic growth and social 3 progress (Brandão Santana et al. 2015; Heeks and Stanforth 2015). The economic benefit derived from 4 increased production and consumption of goods as well as services create higher demands for improved 5 technologies (Gossart 2015). Since the Industrial Revolution, however, and notwithstanding the 6 benefits, this production and consumption trend and the technological changes associated to it have also 7 come at the cost of long-term damage to the life support systems of our planet (Alarcón and Vos 2015; 8 Steffen et al. 2015). The significance of the such impacts depends on the technology, but also on the 9 intrinsic characteristics of the country or region analysed (Brandão Santana et al. 2015). 10 Other chapters in this volume have discussed technological change in various ways, including as a 11 framing issue (chapter 1), in the context of specific sectors (chapters 6-11), for specific purposes 12 (chapter 12) and as a matter of policy, international cooperation and finance (chapters 13-15). Chapter 13 2 discusses past trends in technological change and chapters 3 and 4 discuss it in the context of future 14 modelling. In general, implicitly or explicitly, technological change is assigned an important role in 15 climate change mitigation and achieving sustainable development (Thacker et al. 2019), also in past 16 IPCC reports (IPCC 2014, 2018a). Chapter 16 describes how a well-established innovation system at a 17 national level, guided by well-designed policies, can contribute to achieve mitigation and adaptation 18 targets along with broader sustainable development goals avoiding, in the process, undesired 19 consequences of technological changes. 20 The environmental impacts of social and economic activities, including emissions of GHG, are greatly 21 influenced by the rate and direction of technological changes (Jaffe et al. 2000). Technological changes 22 usually designed and used to increase productivity and reduce the use of natural resources can lead to 23 increasing production and consumption of goods and services through different rebound effects that 24 diminish the potential benefits of reducing the pressure on the environment (Grübler 1998; Kemp and 25 Soete 1990; Gossart 2015; Sorrell 2007; Barker et al. 2009). 26 Those environmental impacts depend not only on which technologies are used, but also on how they 27 are used (Grübler et al. 1999a). Technological change is not exogenous to social and economic systems; 28 technologies are not conceived, selected, and applied autonomously (Grubler et al. 2018). Underlying 29 driving forces of the problem, such as more resource intensive lifestyles and larger populations 30 (Hertwich and Peters 2009; UNEP 2014), remain largely unchallenged. Comprehensive knowledge of 31 the direct and indirect effects of technological changes on physical and social systems could improve 32 decision-making, also in those cases where technological change mitigates environmental impacts. 33 A sustainable global future for people and nature requires rapid and transformative societal change by 34 integrating technical, governance (including participation), financial and societal aspects of the 35 solutions to be implemented (Pörtner et al. 2021; Sachs et al. 2019). A growing body of interdisciplinary 36 research from around the world can inform implementation of adaptive solutions that address the 37 benefits and drawbacks of linkages in social-ecological complexity, including externalities and rebound 38 effects from innovation and technological transformation (Pörtner et al. 2021; Balvanera et al. 2017). 39 Technological change and transitional knowledge can reinforce each other. The value of traditional 40 wisdom and its technological practices provide examples of sustainable and adaptive systems that could 41 potentially adapt to and mitigate climate change (Singh et al. 2020; Kuoljok 2019). Peasants and 42 traditional farmers have been able to respond well to climate changes through their wisdom and 43 traditional practices (Nicholls and Alteri 2013). The integration of the traditional wisdom with new 44 technologies can offer new and effective solutions (Galloway McLean 2010). Do Not Cite, Quote or Distribute 16-8 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Achieving climate change mitigation and other sustainable development goals thus also requires rapid 2 diffusion of knowledge and technological innovations. However, these are hampered by various 3 barriers, some of which are illustrated in Table 16.1 (Markard et al. 2020). 4 5 Table 16.1 Overview of Challenges to Accelerated Diffusion of Technological Innovations. Based on 6 (Markard et al. 2020) Challenges Description Examples Innovations in Since entire systems are changing, also changes Decentralization of electricity supply whole systems in system architecture are needed, which may and integration of variable sources. not keep pace. Interaction between Simultaneous, accelerating changes multiple Electrification of transport, heating multiple systems systems or sectors, vying for the same resources and industry all using the same and subsystems and showing other interactions. renewable electricity source. Industry decline Decline of existing industries and businesses Traditional car industry leading to and incumbent can lead to incumbents slowing down change, facture closures, demise of coal resistance and resistance from e.g. unions or workers. mining and coal-fired power generation leading to local job loss. Consumers and Consumers need to change practices and Less car ownership in a sharing social practices demand patterns. economy, trip planning for public and non-motorised transport, fuelling practices in electric driving. Coordination in Increasing complexity of governance requires Multi-level governance between governance and coordination between multiple levels of European Commission and member policy government and a multitude of actors relevant states in Energy Union package. to the transition, e.g. communities, financial institutions, private sector. 7 The literature on how, in a systemic way, the barriers to sustainability transition can be overcome in 8 various circumstances has been growing rapidly over the past decades. A central element is that national 9 systems of innovation can help achieving both climate change and sustainable development goals, by 10 integrating new ideas, devices, resources, new and traditional knowledge and technological changes for 11 more effective and adaptive solutions (Lundvall 1992). At the organizational level, innovation is seen 12 as a process that can bring value by means of creating more effective products, services, processes, 13 technologies, policies and business models that are applicable to commercial, business, financial and 14 even societal or political organizations (Brooks 1980; Arthur 2009). 15 The literature refers to the terms “technology push,” “market pull,” “regulatory push-pull,” and “firm 16 specific factors” as drivers for innovation, mostly to inform policymakers (Zubeltzu-Jaka et al. 2018). 17 There has also been growing interest in social drivers, motivated by the recognition of social issues, 18 such as unemployment and public health, linked to the deployment of innovative low-carbon 19 technologies (Altantsetseg et al. 2020). Policy and social factors and the diverse trajectories of 20 innovation are influenced by regional and national conditions (Tariq et al. 2017), and such local needs 21 and purposes need to be considered in crafting international policies aimed at fostering the global 22 transition towards increased sustainability (Caravella and Crespi 2020). From this standpoint, a 23 multidimensional, multi-actor systemic innovation approach would be needed to enhance global 24 innovation diffusion (de Jesus and Mendonça 2018), especially if this is to lead to overall sustainability 25 improvements rather than resulting in new sustainability challenges. Do Not Cite, Quote or Distribute 16-9 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Policies to mitigate climate change do not always take into account the effects of mitigation 2 technologies on other environmental and social challenges (Arvesen et al. 2011). Policies also often 3 disregard the strong linkages between technological innovation and social innovation; the latter 4 understood as the use of soft technologies that brings about transformation through establishing new 5 institutions, new practices, and new models to create a positive societal impact characterized by 6 collaboration that crosses traditional roles and boundaries, between citizens, civil society, the state, and 7 the private sector (Reynolds et al. 2017). Market forces do not provide sufficient incentives for 8 investment in development or diffusion of technologies, leaving a role for public policy to create the 9 conditions to assure a systemic innovation approach (Popp 2010; Popp and Newell 2012). Moreover, 10 public action is more than just addressing market failure, it is an unalienable element of an innovation 11 system (Mazzucato 2013). 12 Coupling technological innovation with sustainable development and the SDGs would need to address 13 overall social, environmental, and economic consequences, given that public policy is intertwined with 14 innovation, technological changes and other factors in a complex manner. Chapter 16 is organized in 15 the following manner to provide an overview of innovation and technology development and transfer 16 for climate change and sustainable development. 17 Section 16.2 discusses drivers of innovation process, including macro factors that can redirect 18 technological change towards low-carbon options. Representations of these drivers in mathematical and 19 statistical models allow for explaining the past and constructing projections of future technological 20 change. They also integrate the analysis of drivers and consequences of technological change within 21 economic-energy-economy (or integrated assessment) models (see Chapter 3). The section also 22 describes the different phases of innovation and metrics, such as the widely used but also criticized 23 technology readiness levels (TRLs). 24 Section 16.3 discusses innovation as a systemic process based on recent literature. While the innovation 25 process is often stylized as a linear process, innovation is now predominantly seen as a systemic process 26 in that it is a result of actions by, and interactions among, a large set of actors, whose activities are 27 shaped by, and shape, the context in which they operate and the user group with which they are 28 engaging. 29 Section 16.4 presents innovation and technology policy, including technology push (e.g., publicly 30 funded R&D) and demand-pull (e.g., governmental procurement programmes) instruments that 31 addresses potential market failures related to innovation and technology diffusion. The section also 32 assesses the cost-effectiveness and other policy assessment criteria introduced in Chapter 13 of 33 innovation policies. 34 In section 16.5, the chapter assesses the role of international cooperation in technology development 35 and transfer, in particular the mechanisms established under the UNFCCC, but also other international 36 initiatives for technology cooperation. The discussion on international cooperation includes information 37 exchange, research, development and demonstration cooperation, access to financial instruments, 38 intellectual property rights, as well as promotion of domestic capacities and capacity building. 39 Section 16.6 describes the role of technology in sustainable development, including unintended effects 40 of technological changes, and synthesizes the chapter. Finally, section 16.7 discusses gaps in knowledge 41 emerging from this chapter. 42 43 Do Not Cite, Quote or Distribute 16-10 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 16.2 Elements, drivers and modelling of technology innovation 2 Models of the innovation process, its drivers and incentives provide a tool for technology assessment, 3 constructing projections of technological change and identifying which macro conditions facilitate 4 development of low-carbon technologies. The distinction between stages of innovation process allows 5 to assess technology readiness (Section 16.2.1). Qualitative and quantitative analysis of main elements 6 underpinning innovation - R&D, learning-by-doing, spillovers allow for explanation of past and project 7 future technological change (Section 16.2.2). In addition, general purpose technologies can play a role 8 in climate change mitigation. 9 In the context of mitigation pathways, the feasibility of any emission reduction targets depends on the 10 ability to promote innovation in low- and zero-carbon technologies, as opposed to any other technology. 11 For this reason, the section reviews the literature of the levers influencing the direction of technological 12 change in favour of low- and zero-carbon technologies (Section 16.2.3). Moreover, representation of 13 drivers in mathematical and statistical models from section 16.2.2 allows integrating its analysis with 14 economic and climate effects within IAMs, hence permitting more precise modelling of decarbonisation 15 pathways (Section 16.2.4). 16 In addition to technological innovation, other innovation approaches are relevant in the context of 17 climate mitigation and more broadly sustainable development (Section 16.6). Frugal innovations, i.e. 18 “good enough” innovations that fulfil the needs of non-affluent consumers mostly in developing 19 countries (Hossain 2018), are characterized by low costs, concentration on core functionalities, and 20 optimised performance level (Weyrauch and Herstatt 2016) and are hence often associated with 21 (ecological and social) sustainability (Albert 2019). Grassroots innovations are products, services and 22 processes developed to address specific local challenges and opportunities, and which can generate 23 novel, bottom-up solutions responding to local situations, interests and values. (Dana et al. 2021; 24 Pellicer-Sifres et al. 2018). 25 26 16.2.1 Stages of the innovation process 27 The innovation cycle is commonly thought of as having three distinct innovation phases on the path 28 between basic research and commercial application: Research and Development (R&D), demonstration, 29 and deployment and diffusion (IPCC 2007). Each of these phases differs with respect to the kind of 30 activity carried out, the type of actors involved and their role, financing needs and the associated risks 31 and uncertainties. All phases involve a process of trial and error, and failure is common; the share of 32 innovation that successfully reaches the deployment phase is small. The path occurring between basic 33 research to commercialization is not linear (see also Section 16.3); it often requires a long time and is 34 characterized by significant bottlenecks and roadblocks. Furthermore, technologies may regress 35 backwards in the innovation cycle, rather than move forward (Skea et al. 2019). Successfully passing 36 from each stage to the next one in the innovation cycle requires overcoming “valleys of deaths” 37 (Auerswald and Branscomb 2003; UNFCCC 2017), most notably the demonstration phase (Frank et al. 38 1996; Weyant 2011; Nemet et al. 2018). Over time, new and improved technologies are discovered; 39 this often makes the dominant technology obsolete, but this is not discussed here further. 40 Table 16.2 summarizes the different innovation stages and main funding actors, and maps phases into 41 the technology readiness levels (TRLs) discussed in Section 16.2.1.4. 42 Do Not Cite, Quote or Distribute 16-11 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Table 16.2 Stages of the innovation process (16.2.1) mapped onto Technology Readiness Levels (16.2.1.4) Stage Main funding Phases Related TRL actors Governments Basic research 1 – Initial idea (basic principles defined) Research and development Firms Applied 2 – Application formulated (technology concept and research and application of solution formulated) technology 3 – Concept needs validation (solutions need to be development prototyped and applied 4 – Early prototype (prototype proven in test conditions) 5 – Full prototype at scale (components proven in conditions to be deployed) Governments Experimental 6 – Full prototype at scale (prototype proven at scale in pilot project or conditions to be deployed) Firms full scale testing Demonstration 7 – Pre-commercial demonstration (solutions working in Venture Capital expected conditions) Angel investors 8 – First-of-a-kind commercial (commercial demonstration, full scale deployment in final form) 9 – Commercial operation in early environment (solution is commercial available, needs evolutionary improvement to Firms Commercializat stay competitive) ion and scale up Private equity 10 – Integration needed at scale (solution is commercial and Deployment and diffusion (business) competitive but needs further integration efforts) Commercial banks 11 – Proof of stability reached (Predictable growth) Mutual funds International Transfer organizations and financial institutions NGOs 2 Adapted from: Auerswald and Branscomb (2003), Technology Executive Committee (2017), IEA (IEA 2020a) 3 4 16.2.1.1 Research and Development 5 This phase of the innovation process focuses on generating knowledge or solving particular problems 6 by creating a combination of artefacts that is intended to perform a particular function, or to achieve a 7 specific goal. R&D activities comprise basic research, applied research and technology development. 8 Basic research is experimental or theoretical work undertaken primarily to acquire new knowledge of 9 the underlying foundations of phenomena and observable facts, without any particular application or 10 use in view. Applied research is original investigation undertaken in order to acquire new knowledge, 11 primarily directed towards a specific, practical aim or objective (OECD 2015a). Importantly, R&D 12 activities can be incremental, i.e. focused on addressing a specific need by marginally improving an 13 already existing technology, or radical, representing a paradigm shift, promoted by new opportunities 14 arising with the accumulation of new knowledge (Mendonça et al. 2018). Technology development, 15 often leading to prototyping, consists of generating a working model of the technology that is usable in 16 the real world, proving the usability and customer desirability of the technology and giving an idea of 17 its design, features and functioning (OECD 2015a). These early stages of technological innovation are Do Not Cite, Quote or Distribute 16-12 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 referred to as “formative phase”, during which the conditions are shaped for a technology to emerge 2 and become established in the market (Wilson and Grubler 2013) and the constitutive elements of the 3 innovation system emerging around a particular technology are set up (see Section 16.3)(Bento et al. 4 2018; Bento and Wilson 2016). 5 The outcomes of R&D are uncertain: the amount of knowledge that will result from any given research 6 project or investment is unknown ex ante (Rosenberg 1998). This risk to funders (Goldstein and 7 Kearney 2020) translates into underinvestment in R&D due to low appropriability (Sagar and Majumdar 8 2014; Weyant 2011). In the case of climate mitigation technologies, low innovation incentives for the 9 private sector also result from a negative environmental externality (Jaffe et al. 2005). Furthermore, in 10 absence of stringent climate policies and targets, incumbent fossil-based energy technologies are 11 characterized by lower financing risk, are heavily subsidized (Kotchen 2021; Davis 2014) and 12 depreciate slowly (see Section 16.2.3) (Nanda et al. 2016; Semieniuk et al. 2021; Arrow 1962a). In this 13 context public research funding therefore plays a key role in supporting high-risk R&D both in 14 developed and developing economies: it can provide patient and steady funding not tied to short-term 15 investment returns (see Section 16.4) (Anadon et al. 2014; Mazzucato 2015a; Howell 2017; Zhang et 16 al. 2019; Anadón et al. 2017; Chan and Diaz Anadon 2016; Kammen and Nemet 2007). Public policies 17 also play a role increasing private incentives in energy research and development funding (Nemet 2013). 18 R&D statistics are an important indicator of innovation and are collected following the rules of the 19 Frascati Manual (OECD 2015b) (Section 16.3.3, Box 16.3, Table 16.7). 20 16.2.1.2 Demonstration 21 Demonstration is carried out through pilot projects or large-scale testing in the real world. Successfully 22 demonstrating a technology shows its utility and that it is able to achieve its intended purpose and, 23 consequently, that the risk of failure is reduced (i.e. that it has market potential) (Hellsmark et al. 2016). 24 Demonstration project are an important step to promote the deployment of low-carbon energy and 25 industrial technologies in the context of the transition. government funding often plays a large role in 26 energy technology demonstration projects because scaling up hardware energy technologies is 27 expensive and risky (Brown and Hendry 2009; Hellsmark et al. 2016). Governments’ engagement in 28 low-carbon technology demonstration also signals support for business willing to take the investment 29 risk (Mazzucato 2016). Venture capital, traditionally not tailored for energy investment, can play an 30 increasingly important role also thanks to the incentives (e.g. through de-risking) provided by public 31 funding and policies (Gaddy et al. 2017; IEA 2017a). 32 16.2.1.3 Deployment and diffusion 33 Deployment entails producing a technology at large scale and scaling up its adoption use across 34 individual firms or households in a given market, and across different markets (Jaffe 2015). In the 35 context of climate change mitigation and adaptation technologies, the purposeful diffusion to 36 developing countries, is referred to as “technology transfer”. Most recently, the term " innovation 37 cooperation" has been proposed to indicate that technologies needs to be co-developed and adapted to 38 local contexts (Pandey et al. 2021). Innovation cooperation is an important component of stringent 39 mitigation strategies as well as international agreements (see Section 16.5). 40 Diffusion is often sluggish due to lock-in of dominant technologies (Liebowitz and Margolis 1995; 41 Unruh 2000; Ivanova et al. 2018), as well as the time needed to diffuse information about the 42 technologies, heterogeneity among adopters, the incentive to wait until costs fall even further, the 43 presence of behavioural and institutional barriers and the uncertainty surrounding mitigation policies 44 and long-term commitments to climate targets (Corey 2014; Haelg et al. 2018; Gillingham and Sweeney 45 2012; Jaffe 2015). In addition, novel technology has been hindered by the actions of powerful 46 incumbents who accrue economic and political advantages over time, as in the case of renewable 47 energy generation (Unruh 2002; Supran and Oreskes 2017; Hoppmann et al. 2019). Do Not Cite, Quote or Distribute 16-13 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Technologies have been shown to penetrate the market with a gradual non-linear process in a 2 characteristic logistic (S-shaped) curve (Rogers 2003; Grübler 1996). The time needed to reach 3 widespread adoption varies greatly across technologies relevant for adaptation and mitigation (Gross et 4 al. 2018); in the case of energy technologies, the time needed for technologies to get from a 10 to 90% 5 market share of saturation ranges between 5 to over 70 (Wilson 2012). Investment in commercialization 6 of low-emission technology is largely provided by private financiers; however, governments play a key 7 role in ensuring incentives through supportive policies, including R&D expenditures providing signals 8 to private investors (Haelg et al. 2018), pricing carbon dioxide emissions, public procurement, 9 technology standards, information diffusion and the regulation for end-life cycle treatment of products 10 (Cross and Murray 2018) (see Section 16.4). 11 16.2.1.4 Technology Readiness Levels 12 Technology Readiness Levels (TRLs) are a categorization that enables consistent, uniform discussions 13 of technical maturity across different types of technology. They were developed by NASA in the 1970s 14 (Mankins 2009, 1995) and originally used to describe the readiness of components forming part of a 15 technological system. Over time, more classifications of TRLs have been introduced, notably the one 16 used by the EU. Most recently, the IEA extended previous classifications to include the later stages of 17 the innovation process (IEA 2020b) and applied it to compare the market readiness of clean energy 18 technologies and their components (OECD 2015a; IEA 2020b). TRLs are currently widely used by 19 engineers, business people, research funders and investors, often to assess the readiness of whole 20 technologies rather than single components. To determine a TRL for a given technology, a Technology 21 Readiness Assessment (TRA) is carried out to examine programme concepts, technology requirements, 22 and demonstrated technology capabilities. In the most recent version of the IEA (IEA 2020b), TRLs 23 range from 1 to 11, with 11 indicating the most mature (see Table 16.2). 24 The purpose of TRLs is to support decision making. They are applied to avoid the premature application 25 of technologies, which would lead to increased costs and project schedule extension (US Department 26 of Energy 2011). They are used for risk management, and can also be used to make decisions regarding 27 technology funding and to support the management of the R&D process within a given organization or 28 country (De Rose et al. 2017). 29 In practice, the usefulness of TRLs is limited by several factors. These include limited applicability in 30 complex technologies or systems, the fact that they do not define obsolescence, nor account for 31 manufacturability, commercialization or the readiness of organizations to implement innovations 32 (European Association of Research Technology Organisations 2014) and do not consider any type of 33 technology-system mismatch or the relevance of the products’ operation environment to the system 34 under consideration (Mankins 2009). Many of these limitations can be eased by using TRLs in 35 combination with other indicators such as System Readiness Levels and other economic indicators on, 36 for example, investments and returns (IEA 2020b). 37 38 16.2.2 Sources of technological change 39 The speed of technological change could be explained with the key drivers of innovations process: R&D 40 effort, learning-by-doing and spillover effects. In addition, new innovations are sometimes enabled by 41 the development of general purpose technologies, such as digitalization. 42 16.2.2.1 Learning-by-doing and research and development 43 Learning by doing and R&D effort are two factors commonly used by the literature to explain past and 44 project future speed of technological change (Klaassen et al. 2005; Mayer et al. 2012; Bettencourt et al. 45 2013). Learning-by-doing is the interaction of workers with new machines or processes that allows 46 them to use them more efficiently (Arrow 1962b). R&D effort is dedicated to looking for new solutions Do Not Cite, Quote or Distribute 16-14 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 (e.g. blueprints) that could increase the efficiency of existing production methods or result in entirely 2 new methods, products or services (see section 16.2.1.1). 3 Learning-by-doing and research and development are interdependent. Young (1993) postulates that 4 learning-by-doing cannot continue forever without R&D because it is bounded by an upper physical 5 productivity limit of an existing technology. Research and development can shift this limit because it 6 allows replacing the existing technology with a new one. On other hand, incentives to invest in R&D 7 depend on future cost of manufacturing, which in turn depend on the scale of learning-by-doing. The 8 empirical evidence for virtuous circle between costs reduction, market growth and R&D were found in 9 the case of PV market ((Watanabe et al. 2000); see also Box 16.4), but could also lead to path 10 dependency and lock-in (Erickson et al. 2015). Section 16.4.4 and Chapter 13 Section 13.7.3.1. discuss 11 how simultaneous use of technology push and pull policies could amplify effects of research and 12 learning. 13 The benefits of R&D and learning-by-doing are larger at the economy level than at the firms level 14 (Romer 1990; Arrow 1962b). As a result, the market, left to its own, tends to generate less investment 15 than socially optimal. For instance, if the cost of a technology is too high before a large amount of 16 learning-by-doing has occurred, there is a risk that it will not be adopted by the market even if it is 17 economically advantageous for the society. Indeed, initially new technologies are often expensive and 18 cannot compete with the incumbent technologies (Cowan 1990). Large numbers of adopters could lower 19 this cost via learning-by-doing to a level sufficient to beat the incumbent technology (Gruebler et al. 20 2012). However, firms could hesitate to be the first adopter and bear the high cost (Isoard and Soria 21 2001). If this disadvantage overwhelms the advantages of being a first mover1 and if adopters are not 22 able to coordinate, it will lead to situation of a lock-in (Gruebler et al. 2012). 23 The failure of markets to deliver the size of R&D investment and learning-by-doing that would be 24 socially optimal is one of the justifications of government intervention. Policies to address these market 25 failures can be categorized as technology push and demand pull policies. The role of these policies is 26 explained in Table 16.3. Section 16.4 discusses individual policy instruments in greater detail. 27 28 Table 16.3 Categories of policies and interventions accelerating technological changes, the factors 29 promoting them and slowing them down, illustrated with examples What it refers to: What promotes What slows down Examples technological technological change change Technology Support the R&D, funding and Inadequate supply Japan’s Project Sunshine, the Push creation of new performance of of trained US Project Independence in knowledge to early scientists and the 1970s. Breakthrough make it easier to demonstrations engineers (Popp Energy Coalition and Mission invest in (Brown and Hendry and Newell 2012); Innovation, respectively innovation 2009; Hellsmark et gap with demand private- and public-sector al. 2016) pull (Grübler et al. international collaborations to 1999b). respectively focus energy innovation and double energy R&D, both initiated concurrently with the Paris Agreement in 2015 (Sanchez and Sivaram 2017). FOOTNOTE1 see e.g. (Spence 1981) and (Bhattacharya 1984) for discussion of first-mover advantages Do Not Cite, Quote or Distribute 16-15 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII Demand Instruments Enlarging potential Willingness of Subsidies for wind power Pull creating market markets, increasing consumers to California, the German feed-in opportunities. adoption of new accept new tariff for PV, quotas for electric fuels and mitigation technology. Policy vehicles in China (Wang et al. technology. and political 2017a) and Norway volatility can deter (Pereirinha et al. 2018) Digital innovations investment. Biofuels (Brazil); Social Social innovation innovation with Wind Energy and awareness (Denmark, Germany) 1 The size of learning-by-doing effect is quantified in literature using learning rates i.e. estimates of 2 negative correlation between costs and size of deployment of technologies. The results from this 3 literature include estimates for energy technologies (McDonald and Schrattenholzer 2001), electricity 4 generation technologies (Rubin et al. 2015; Samadi 2018), for storage (Schmidt 2017), for end-of-pipe 5 control (Kang et al. 2020) and for energy demand and energy supply technologies (Weiss et al. 2010). 6 Meta-analyses find learning rates vary across technologies, within technologies and over time (Wei et 7 al. 2017; Nemet 2009a; Rubin et al. 2015). Moreover, different components of one technology have 8 different learning rates (Elshurafa et al. 2018). Central tendencies are around 20% cost reduction for 9 each doubling of deployment (McDonald and Schrattenholzer 2001). 10 Studies of correlation between cumulative deployment of technologies and costs are not sufficiently 11 precise to disentangle the causal effect of increase in deployment from the causal effects of research 12 and development and other factors (Nemet 2006). Numerous subsequent studies attempted to, amongst 13 others, separate the effect of learning-by-doing and research and development (Klaassen et al. 2005; 14 Mayer et al. 2012; Bettencourt et al. 2013), economies of scale (Arce 2014), and knowledge spillovers 15 (Nemet 2012). Once those other factors are accounted for, some empirical studies find that the role of 16 learning-by-doing in driving down the costs becomes minor (Kavlak et al. 2018; Nemet 2006). In 17 addition the relation could reflect reverse causality: increase in deployment could be an effect (and not 18 a cause) of a drop in price (Witajewski-Baltvilks et al. 2015; Nordhaus 2014). Nevertheless, in some 19 applications, learning curves can be a useful proxy and heuristic (Nagy et al. 2013). 20 The negative relation between costs and experience is a reason to invest in a narrow set of technologies; 21 the uncertainty regarding the parameters of this relation is the reason to invest in wider ranges of 22 technologies (Way et al. 2019; Fleming and Sorenson 2001). Concentrating investment in narrow sets 23 of technologies (specialization) enables fast accumulation of experience for these technologies and large 24 cost reductions. However, when the potency of technology is uncertain, one does not know which 25 technology is truly optimal in the long-run. The narrower the set the higher the risk that the optimal 26 technology will not be supported, and hence will not benefit from learning-by-doing. Widening the set 27 of supported technologies would reduce this risk (Way et al. 2019). Uncertainty is present because noise 28 in historical data hides the true value of learning rates as well as because of unanticipated future shocks 29 to technology costs (Lafond et al. 2018). Ignoring uncertainty in integrated assessment models implies 30 that these model results are biased towards supporting narrow set of technologies neglecting the benefits 31 of decreasing risk trough diversification (Sawulski and Witajewski-Baltvilks 2020). 32 16.2.2.2 Knowledge spillovers 33 Knowledge spillovers drive continuous technological change (Rivera-Batiz and Romer 1991; Romer 34 1990) and are for that reason relevant to climate technologies as well as incumbent, carbon-intensive 35 technologies. Knowledge embedded in innovations by one innovator give an opportunity for others to 36 create new innovations and increase the knowledge stock even further. The constant growth of 37 knowledge stock through spillovers translates into constant growth of productivity and cost reduction. Do Not Cite, Quote or Distribute 16-16 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 By allowing for experimenting with existing knowledge and combining different technologies, 2 knowledge spillovers can result in the emergence of novel technological solutions, which has been 3 referred to as recombinant innovation (Weitzman 1998; Olsson and Frey 2002; Tsur and Zemel 2007; 4 Arthur 2009; Fleming and Sorenson 2001). Recombinant innovations speed up technological change 5 by combining different technological solutions, and make things happen that would be impossible with 6 only incremental innovations (Safarzyńska and van den Bergh 2010; van den Bergh 2008; Frenken et 7 al. 2012). It has been shown that 77% of all patents granted between 1790 and 2010 in the US are coded 8 by a combination of at least two technology codes (Youn et al. 2015). Spillovers related to energy and 9 low-carbon technologies has been documented by a number of empirical studies (high confidence) 10 (Popp 2002; Aghion et al. 2013; Witajewski-Baltvilks et al. 2017; Verdolini and Galeotti 2011; Conti 11 et al. 2018). The presence of spillovers can have both positive and negative impacts on climate change 12 mitigation (high confidence). 13 The spillover effect associated with innovation in carbon-intensive technologies may lead to lock-in of 14 fossil-fuel technologies. Continuous technological change of carbon-intensive industry raises the bar 15 for clean technologies: a larger drop in clean technologies’ cost is necessary to become competitive 16 (Acemoglu et al. 2012; Aghion et al. 2013). The implication is that delaying climate policy increases 17 the cost of that policy (Aghion 2019). 18 On the other hand, the spillover effect associated with innovation in low-emission technologies increase 19 the potency of climate policy (Aghion 2019). For instance, a policy that encourages clean innovation 20 leads to accumulation of knowledge in clean industry which, through spillover effect encourages further 21 innovation in clean industries. Once the stock of knowledge is sufficiently large, the value of clean 22 industries will be so high, that technology firms will invest there even without policy incentives. Once 23 this point is reached, the policy intervention can be discontinued (Acemoglu et al. 2012). 24 In addition, the presence of spillovers implies that a unilateral effort to reduce emissions in one region 25 could reduce emissions in other regions (medium confidence) (Gerlagh and Kuik 2014; Golombek and 26 Hoel 2004). For instance, in the presence of spillovers, a carbon tax that incentivises clean technological 27 change increases the competitiveness of clean technologies not only locally, but also abroad. The size 28 of this effect depends on the size of spillovers. If they are sufficiently strong, the reduction of emissions 29 abroad due to clean technological change could be larger than the increase of emissions due to carbon 30 leakage (Gerlagh and Kuik 2014). Different types of carbon leakage are discussed in Chapter 13, 31 Section 13.7.1 and other consequences of spillovers for the design of policy are discussed in Chapter 32 13, Section 13.7.3. 33 16.2.2.3 General purpose technologies and digitalization 34 General purpose technologies (GPTs) provide solutions that could be applied across sectors and 35 industries (Goldfarb 2011) by creating technological platforms for a growing number of interrelated 36 innovations. Examples of GPTs relevant to climate change mitigation are hydrogen and fuel cell 37 technology, which may find applications in transport, industry and distributed generation (Hanley et al. 38 2018), and nanotechnology which played a significant role in advancement of all the different types of 39 renewable energy options (Hussein 2015). Assessing the environmental, social and economic 40 implications of such technologies, including increased emissions through energy use, is challenging 41 (see Chapter 5, Section 5.3.4.1 and Cross-Chapter Box 11 below). 42 Several GPTs relevant for climate mitigation and adaptation emerged as a result digitalization, namely 43 the adoption or increase in use of information and communication technologies (ICTs) by citizens, 44 organizations, industries or countries and the associated restructuring of several domains of social life 45 and of the economy around digital technologies and infrastructures (IEA 2017b; Brennen and Kreiss 46 2016). The digital revolution is underpinned by innovation in key technologies, e.g. ubiquitous 47 connected consumer devices such as mobile phones (Grubler et al. 2018), rapid expansions of global 48 internet infrastructure and access (World Bank 2014), and steep cost reductions and performance Do Not Cite, Quote or Distribute 16-17 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 improvements in computing devices, sensors, and digital communication technologies (Verma et al. 2 2020). The increasing the pace at which the physical and digital worlds are converging increase the 3 relevance of disruptive digitalization in the context of climate mitigation and sustainability challenges 4 (European Commission 2020) (see Cross-chapter Box 11 in this chapter and Chapter 4, Section 4.4.1). 5 Digital technologies require energy, but increase efficiency, potentially offering technology-specific 6 GHG emission savings; they also have larger system wide impacts (Kaack et al. 2021). In industrial 7 sectors, robotization, smart manufacturing (SM), internet of things (IoT), artificial intelligence (AI), 8 and additive manufacturing (AM or 3D printing), have the potential to reduce material demand and 9 promote energy management (Chapter 11, Section 11.3.4.2). Smart mobility is changing transport 10 demand and efficiency (Chapter 10, Section 10.2.3). Smart devices in buildings, the deployment of 11 smart grids and the provision of renewable energy increase the role of demand-side management 12 (Serrenho and Bertoldi 2019) (Chapter 9, Sections 9.4 and 9.5), and support the shift away from asset 13 redundancy (Chapter 6, Section 6.4.3). Digital solutions are equally important on the supply side, for 14 example by accelerating innovation with simulations and deep learning (Rolnick et al. 2021) or realizing 15 flexible and decentralized opportunities through energy-as-a-service concepts and particularly with 16 Pay-As-You-Go (Chapter 15, Box 15.8, Table 1). 17 Yet, increased digitalization could give increase energy demand, thus wiping away potential efficiency 18 benefits, unless appropriately governed (IPCC 2018a). Moreover, digital technologies could negatively 19 impact labour demand and increase inequality (Cross-Chapter Box 11 in this chapter). 20 21 START CCB 11 HERE 22 Cross-Chapter Box 11: Digitalization: efficiency potentials and governance considerations 23 Felix Creutzig (Germany), Elena Verdolini (Italy), Paolo Bertoldi (Italy), Luisa F. Cabeza 24 (Spain), María Josefina Figueroa Meza (Venezuela/Denmark), Kirsten Halsnæs (Denmark), Joni 25 Jupesta (Indonesia), Şiir Kilkiş (Turkey), Michael Koenig (Germany), Eric Masanet (the United States 26 of America), Nikola Milojevic-Dupont (France), Joyashree Roy (India/Thailand), Ayyoob Sharifi 27 (Iran/Japan). 28 Digital technologies impact positively and negatively GHG emissions through their own carbon 29 footprint, via technology application for mitigation, and via induced larger social change. Digital 30 technologies also raise broader sustainability concerns due to their use of rare materials and 31 associated waste, and their potential negative impact on inequalities and labour demand. 32 Direct impacts emerge because digital technologies consume large amounts of energy, but also 33 have the potential to steeply increase energy efficiency in all end-use sectors through material 34 input savings and increased coordination (medium evidence, medium agreement) (Horner et al. 35 2016; Jones 2018) (Huang et al. 2016; IEA 2017b). Global energy demand from digital appliances 36 reached 7.14 EJ in 2018 (Chapter 9, Box 9.5), implying higher related carbon emissions. However, a 37 small smart phone offers services previously requiring many different devices (Grubler et al. 2018). 38 Demand for data services is increasing rapidly; quantitative estimates of the growth of associated energy 39 demand range from slow and marginal to rapid and sizeable, depending the efficiency trends of digital 40 technologies (see Chapter 5.3.4.1)(Avgerinou et al. 2017; Stoll et al. 2019; Vranken 2017; Masanet et 41 al. 2020). Renewable energy serves as low-carbon energy provider for the operation of data centre, 42 which in turn can provide waste heat for other purposes. Digital technologies can markedly increase the 43 energy efficiency of mobility and residential and public buildings, especially in the context of systems 44 integration (IEA 2020a). Reduction in energy demand and associated GHG emissions from buildings 45 and industry while maintaining service levels equal is estimated at 5 to 10%, with larger savings 46 possible. Approaches include building energy management systems (BEMS), home energy Do Not Cite, Quote or Distribute 16-18 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 management system (HEMS), demand response and smart charging (Cross-Chapter Box 11, Table 1. 2 Data centres can also play a role in energy system management, e.g., by increasing renewable energy 3 generation through predictive control (Dabbagh et al. 2019), and by helping drive the market for battery 4 storage and fuel cells (Riekstin et al. 2014). Temporal and spatial scheduling of electricity demand can 5 provide about 10 GW in demand response in the European electricity system in 2030 (Koronen et al. 6 2020; Wahlroos et al. 2017, 2018; Laine et al. 2020). 7 However, system-wide effects may endanger energy and GHG emission savings (high evidence, 8 high agreement). Economic growth resulting from higher energy and labour productivities can increase 9 energy demand (Lange et al. 2020) and associated GHG emissions. Importantly, digitalization can also 10 benefit carbon-intensive technologies (Victor 2018). Impacts on GHG emissions are varied in smart 11 and shared mobility systems, as ride hailing increases GHG emissions due to deadheading, whereas 12 shared pooled mobility and shared cycling reduce GHG emissions, as occupancy levels and/or weight 13 per person km transported improve (Chapter 5, Section 5.3). Energy and GHG emissions impacts from 14 the ubiquitous deployment of smart sensors and service optimization applications in smart cities through 15 are insufficiently assessed in the literature (Milojevic-Dupont and Creutzig 2021). Systemic effects have 16 wider boundaries of analysis, including broader environmental impacts (e.g. demand for rare materials, 17 disposal of digital devices). These need to be integrated holistically within policy design (Kunkel and 18 Matthess 2020), but they are difficult to quantify and investigate (Bieser and Hilty 2018). Policies and 19 adequate infrastructures and choice architectures can help manage and contain the negative 20 repercussions of systemic effects (Section 5.4, 5.6, 9.9). 21 22 Cross-chapter Box 11, Table 1. Selected sector approaches for reducing GHG emissions that are 23 supported by new digital technologies. Contributions of digitalization include a) supporting role (+), b) 24 necessary role in mix of tools (++), c) necessary unique contribution (+++), but digitalization may also 25 increase emissions (−). See also chapters 5, 8, 9, and 11. Sector Approach Quantitative Contribution of Systems References evidence digitalization perspective and broader societal impacts Residential Nudges 2-4% reduction + New (Buckley 2020; energy use (feedback, in global appliances Zangheri et al. in combination information, household increase 2019; Khanna with monetary etc.) energy use consumption et al. 2021; incentives, non- possible Nawaz et al. digital information 2020) Smart Shared Reduction for − or ++ Ride hailing (OECD and ITF mobility mobility and shared cycling increases GHG 2020; Zeng et digital and shared Apps together with emissions, al. 2017) feedback pooled big data and especially due (ecodriving) mobility; machine learning to deadheading increase for algorithm key ride hailing/ precondition for ride sourcing; new shared reduction for mobility ecodriving Smart cities Using digital Precise data ++ Efficiency (Milojevic- devices and big about roadway gains are often Dupont and Big data analysis data to make use can reduce compensated Creutzig 2021) necessary for urban transport material by more optimization (Chapter 10, and building intensity and driving and Box 10.1) associated other rebound GHG effects; privacy Do Not Cite, Quote or Distribute 16-19 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII use more emissions by concerns efficient 90%, linked with digital devices in homes Agriculture Precision Very high + ICTs provide The digital (Townsend et agriculture potential for information and divide is al. 2019; through variable-rate technologies which growing fast, Deichmann et sensors and nitrogen enables farmers to especially al. 2016; Soto satellites application, increase yields, between Embodas et al. providing moderate optimize crop modern and 2019; information on potential for management, subsistence Chlingaryan et soil moisture, variable-rate reduce fertilizers farming; al. 2018) temperature, irrigation and pesticides, Privacy and crop growth feed and water; data may erode and livestock increases trust in feed levels, efficiency of technologies labour-intensive tasks Industry Industrial Process, ++ increased Optimization (GeSI 2012; Internet of activity & efficiency in value chains Parida et al. Things (IIoT) functional can reduce 2019; Rolnick optimization ++ 1.3 Gt CO2-eq wasted et al. 2021; increases estimated resources Wang et al. energy and abatement potential 2016) carbon in manufacturing efficiency + promote sustainable business models Load Big data Reduces + Accelerated Facilitate (Akorede et al. management analysis for capacity experimentation in integration of 2010; Hirsch et and battery optimizing intended for material science renewable al. 2018; de storage demand peak demand, with artificial energy sources Sisternes et al. optimization management shifts demand intelligence 2016; Sivaram Improve and using to align with 2018a; Gür ++ / +++ Forecast utilization of flexible load of intermittent 2018; Aghaei and control generation appliances renewable and Alizadeh algorithms for assets with batteries energy 2013; Vázquez- storage and availability System-wide Canteli and dispatch rebound effects Nagy 2019; management possible Voyant et al. 2017) (Chapter 6, Section 6.4) 1 2 Broader societal impacts of digitalization can also influence climate mitigation because of induced 3 demand for consumption goods, impacts on firms’ competitiveness, changes the demand for skills 4 and labour, worsening of inequality – including reduced access to services due to the digital divide 5 – and governance aspects (low evidence, medium agreement) (Chapter 4, Section 4.4, Chapter 5, 6 Sections 5.3, 5.6). Digital technologies expand production possibilities in sectors other than ICTs 7 through robotics, smart manufacturing, and 3D printing, and have major implications on consumption 8 patterns (Matthess and Kunkel 2020). Initial evidence suggests that robots displace routine jobs and 9 certain skills, change the demand for high-skilled and low-skilled workers, and suppress wages 10 (Acemoglu and Restrepo 2019). Digitalization can thus reduce consumers’ liquidity and consumption 11 (Mian et al. 2020) and contribute to global inequality, including across the gender dimension, raising 12 fairness concerns (Kerras et al. 2020; Vassilakopoulou and Hustad 2021). Digital technologies can lead 13 to additional concentration in economic power (e.g. Rikap 2020) and lower competition; open source Do Not Cite, Quote or Distribute 16-20 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 digital technologies can however counter this tendency (e.g. Rotz et al. 2019). Digital technologies play 2 a role in mobilizing citizens for climate and sustainability actions (Westerhoff et al. 2018; Segerberg 3 2017). 4 Whether the digital revolution will be an enabler or a barrier for decarbonization will ultimately 5 depend on the governance of both digital decarbonization pathways and digitalization more in 6 general (medium evidence, high agreement). The understanding of the disruptive potential of the wide 7 range of digital technologies is limited due to their ground-breaking nature, which makes it hard to 8 extrapolate from previous history/experience. Municipal and national entities can make use of digital 9 technologies to manage and govern energy use and GHG emissions in their jurisdiction (Bibri 2019a,b) 10 and break down solution strategies to specific infrastructures, building, and places, relying on remote 11 sensing and mapping data, and contextual (machine-) learning about their use (Milojevic-Dupont and 12 Creutzig 2021). Mobility apps can provide mobility-as-a-service access to cities ensuring due 13 preference to active and healthy modes (see 9.9 for the example of the Finnish city of Lahti). Trusted 14 data governance can promote the implementation of local climate solutions, supported by available big 15 data on infrastructures and environmental quality (Hansen and Porter 2017; Hughes et al. 2020). 16 Governance decisions, such as taxing data, prohibiting surveillance technologies, or releasing data that 17 enable accountability can change digitalization pathways, and thus underlying GHG emission (Hughes 18 et al. 2020). 19 Closing the digital gap in developing countries and rural communities enables an opportunity for 20 leapfrogging (medium evidence, medium agreement). Communication technologies (such as mobile 21 phones) enable participation of rural communities, especially in developing countries, and promote 22 technological leapfrogging, e.g. decentralized renewable energies and smart farming (Ugur and Mitra 23 2017; Foster and Azmeh 2020; Arfanuzzaman 2021). Digital technologies have sector-specific 24 potentials and barriers, and may benefit certain regions/areas/socioeconomic groups more than others. 25 For example, integrated mobility services benefit cities more than rural and peripheral areas (OECD 26 2017). 27 Appropriate mechanisms also need to be designed to govern digitalization as megatrend (medium 28 evidence, high agreement). Digitalization is expected to be a fast process, but this transformation takes 29 place against entrenched individual behaviours, existing infrastructure, the legacy of time frames, 30 vested interest and slow institutional processes and requires trust from consumers, producers and 31 institutions. A core question relates to who controls and manages data created by everyday operations 32 (calls, shopping, weather data, service use, etc.). Regulations that limit or ban the expropriation and 33 exploitation of behavioural data, sourced via smart phones, represent crucial aspects in digitalization 34 pathways, alongside the possibility to create climate movements and political pressure from the civil 35 society. Governance mechanisms need to be developed to ensure that digital technologies such as 36 artificial intelligence take over ethical choices (Craglia et al. 2018; Rahwan et al. 2019). Appropriate 37 governance is necessary for digitalization effectively work in tandem with established mitigation 38 technologies and choice architectures. Consideration of system-wide effects and overall management 39 is essential to avoid run-away effects. Overall governance of digitalization remains a challenge, and 40 will have large-scale repercussions on energy demand and GHG emissions. 41 END CCB 11 HERE 42 16.2.2.4 Explaining past and projecting future technology cost changes 43 Researchers and policymakers alike are interested in using observed empirical patterns of learning to 44 project future reductions in costs of technologies. Studies cutting across a wide range of industrial 45 sectors (not just energy) have tried to relate cost reductions to different functional forms, including cost 46 reductions as a function of time (Moore’s law) and cost reductions as a function of production or Do Not Cite, Quote or Distribute 16-21 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 deployment (Wright’s law, also known as Henderson's law), finding that those two forms perform better 2 than alternatives combining different factors, with costs as a function of production (Wright’s law) 3 performing marginally better (Nagy et al. 2013). A comparison of expert elicitation and model-based 4 forecasts of the future cost of technologies for the energy transition indicates that model-based forecast 5 medians were closer to the average realized values in 2019 (Meng et al. 2021). 6 Recent studies attempt to separate the influence of learning-by-doing (which is a basis of Wright’s law) 7 versus other factors in explaining cost reductions specifically in energy technologies. Some studies 8 explain cost reductions with two factors: cumulative deployment (as proxy for experience) and R&D 9 investment (see the “two factor” learning curve (Klaassen et al. 2005). However, reliable information 10 on public energy R&D investments for developing countries is not systematically collected. Available 11 data for OECD countries cannot be precisely assigned to specific industrial sectors or sub-technologies 12 (Verdolini et al. 2018). Some learning-curve studies take into account that historical variation in 13 technology costs could be explained by variation in key materials and fuel costs (for example steel costs 14 for wind turbines (Qiu and Anadon 2012) silicon costs (Kavlak et al. 2018; Nemet 2006) coal and coal 15 plant construction costs (McNerney et al. 2011). Economies of scale played a significant role in the 16 PV cost reductions since the early 2000s (Yu et al. 2011) (See also Box 16.4), which can also be the 17 case in organic PV technologies (Gambhir et al. 2016; Kavlak et al. 2018). 18 19 16.2.3 Directing technological change 20 Technological change is characterized not only by its speed, but also its direction. The early works that 21 considered the role of technology in economic and productivity growth (Solow 1957; Nelson and Phelps 22 1966) assumed that technology can move forward along only one dimension - every improvement led 23 to an increase in efficiency and increased demand for all factors of production. This view however 24 ignores the potency of technological change to alter the otherwise fixed relation between economic 25 growth and the use of resources. 26 Technological change that saves fossil fuels could decouple economic growth and CO2 emissions 27 (Acemoglu et al. 2014; Hémous 2016; Greaker et al. 2018; Acemoglu et al. 2012). Saving of fossils 28 could be obtained with increasing efficiency of producing alternatives to fossils (Acemoglu et al. 2012, 29 2014). This is the case of oil consumption by combustion engine cars which could be substituted with 30 electric cars (Aghion et al. 2013). If there is no close substitute to dirty resource, then its intensity in 31 production could still be reduced by increasing efficiency of the dirty resource relative to efficiency of 32 other inputs (Hassler et al. 2012; André and Smulders 2014; Witajewski-Baltvilks et al. 2017). For 33 instance, energy efficiency improvement leads to drop in relative demand for energy (Hassler et al. 34 2012; Witajewski-Baltvilks et al. 2017). 35 16.2.3.1 Determinants of technological change direction: prices, market size and government 36 Firms change their choice of technology upon change in prices: when one input (e.g. energy) becomes 37 relatively expensive, firms pick technologies which allow them to economize on that input, according 38 to price-induced technological change theory (Reder and Hicks 1965; Samuelson 1965; Sue Wing 39 2006). For example, an increase in oil price will lead to a choice of fuel-saving technologies. Such 40 response of technological change was evident during the oil-price shocks in the 1970s (Hassler et al. 41 2012). Technological change that is induced by an increase in price of a resource can never lead to an 42 increase in use of that resource. In other words, rebound effects associated induced technological change 43 can never offset the saving effect of that technological change (Antosiewicz and Witajewski-Baltvilks 44 2021). 45 The impact of energy prices on the size of low-carbon technological change is supported by large 46 number of empirical studies (Popp 2019; Grubb and Wieners 2020). Studies document that higher Do Not Cite, Quote or Distribute 16-22 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 energy prices are associated with higher number of low-carbon energy or energy efficiency patents 2 (Noailly and Smeets 2015; Ley et al. 2016; Lin and Chen 2019; Newell et al. 1999; Popp 2002; 3 Witajewski-Baltvilks et al. 2017; Verdolini and Galeotti 2011). Sue Wing (2008) finds that innovation 4 induced by energy prices had a minor impact on the decline in U.S. energy intensity in the last decades 5 of 20th century and that autonomous technological change played a more important role. Several studies 6 explore the impact of a carbon tax on green innovation (see Section 16.4). However, disentangling the 7 effect of policy tools is complex because presence of some policies could distort the functioning of 8 other policies (Böhringer and Rosendahl 2010; Fischer et al. 2017) and because the impact of policies 9 could be lagged in time (Antosiewicz and Witajewski-Baltvilks 2021). 10 The direction of technological change depends also on the market size for dirty technologies relative to 11 the size of other markets (Acemoglu et al. 2014). Due to this dependence, climate and trade policy 12 choices in a single region can alter the direction of technological change at the global level (see Section 13 16.2.3.3). 14 The value of the market for clean technologies is determined not only by a current but also by firm’s 15 expectations of future stream of profits (Alkemade and Suurs 2012; Greaker et al. 2018; Aghion 2019). 16 One implication is that bolstering the credibility and durability of policies related to low-carbon 17 technology is crucial to accelerating technological change and inducing the private sector investment 18 required (Helm et al. 2003), especially in rapidly growing economies of Asia and Africa who are on the 19 brink of making major decisions about the type of infrastructure they build as they grow, develop, and 20 industrialize (Nemet et al. 2017). 21 If governments commit to climate policies, firms expect that the future size of markets for clean 22 technologies will be large and they are eager to redirect research effort towards development of these 23 technologies today. Furthermore the commitment would also incentivise acquiring skills that could 24 further reduce the costs of those technologies (Aghion 2019). However, historical evidence shows that 25 policies related to energy and climate over the long term have tended to change (Nemet et al. 2013; 26 Taylor 2012; Koch et al. 2016). Still, where enhancing policy durability has proven infeasible, multiple 27 uncorrelated potentially overlapping policies can provide sufficient incentives (Nemet 2010). 28 16.2.3.2 Determinants of direction of technological change: financial markets 29 The challenges of investing in innovation in energy when compared to other important areas, such as 30 IT and medicine are also reflected in the trends in venture capital funding. Research found that early- 31 stage investments in clean-tech companies were more likely to fail and returned less capital than 32 comparable investments in software and medical technology (Gaddy et al. 2017), which led to a retreat 33 from investors from hardware technologies required for renewable energy generation and storage to 34 software based technologies and demand-side solutions (Bumpus and Comello 2017). 35 The preference for particular types of investments in renewable energy technologies depends on 36 investors attitude to risk (Mazzucato and Semieniuk 2018). Some investors invest in only one 37 technology, others may spread their investments, or invest predominantly in high-risk technologies. The 38 distribution of different types of investors will affect whether finance goes to support deployment of 39 new high-risk technologies, or diffusion of more mature, less-risky technologies characterized by 40 incremental innovations. The role of finance in directing investment is further discussed in Chapter 15, 41 Section 15.6.2. 42 16.2.3.3 Internationalization of green technological change 43 A unilateral effort to reduce emission (via a combination of climate, industrial and trade policies) in a 44 coalition of regions that are technology leaders will reduce cost of clean technologies, which will induce 45 emission reduction in the countries outside the coalition (Di Maria and Smulders 2005; Di Maria and 46 van der Werf 2008; van den Bijgaart 2017; Golombek and Hoel 2004; Hémous 2016). The literature Do Not Cite, Quote or Distribute 16-23 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 suggests various mechanisms leading to this result. Di Maria and van der Werf (2008) argues that the 2 effort to reduce emission in one region reduces global demand for dirty good. This will redirect global 3 innovation towards clean technologies, leading to drop in cost of clean production in every region. 4 The model in Hemous (2016) predict that the coalition could induce acceleration of clean technological 5 change with a mix of carbon tax, clean R&D subsidies and trade policies in that region leading to 6 reduction of cost of clean production inside the coalition. Export of goods produced with clean 7 technologies to a region outside the coalition reduces demand for dirty good in that region. In the model 8 by van den Bijgaart (2017) local advancements of clean technologies by a coalition with strong R&D 9 potential are imitated outside the coalition. Furthermore, advancements of clean technologies will 10 incentivise future clean R&D outside the coalition due to intertemporal knowledge spillovers. In 11 Golombek and Hoel (2004) increase in environmental concern in one region increases abatement R&D 12 in that region. Part of this knowledge spills over to other regions, increasing their incentive to increase 13 abatement too, providing the latter regions did not invest in abatement before. 14 However, this chain breaks if the regions that are behind technological frontier (i.e., technological 15 followers) are not able to absorb the solutions developed by regions at the frontier. New technologies 16 might fail due to deficiencies of political, commercial, industrial, and financial institutions, which we 17 list in table 16.4. For instance, countries might not benefit fully from international knowledge spillover 18 due to insufficient domestic R&D investment, since local knowledge is needed to determine the 19 appropriateness of technologies for the local market, adapting them, installing and using effectively 20 (Gruebler et al. 2012). From the policy perspective this implies that simple transfer of technologies 21 could be insufficient to guarantee adoption of new technologies (Gruebler et al. 2012). 22 23 Table 16.4 Examples of institutional deficiencies preventing deployment of new technologies in countries 24 behind technological frontier. Institutions Examples of deficiencies Literature reference Industrial Inability to benefit fully from international (Mancusi 2008; Unel 2008; Gruebler knowledge spillover due to insufficient domestic et al. 2012) R&D investment Commercial Insufficient experience with the organization and (Abramovitz 1986; Aghion et al. management of large-scale enterprise 2005) Political Vested interests and customary relations among (Olson 1982; Abramovitz 1986). firms and between employers and employees Financial Financial markets incapable of mobilizing capital for (Abramovitz 1986; Aghion et al. individual firms at large scale 2005) 25 26 Research relying on patent citations has indicated that Foreign Direct Investment (FDI) is a mechanism 27 for firms to both contribute to the recipient country’s innovation output as well as benefitting from the 28 recipient country both in industrialized countries (Branstetter 2006) and in developing countries 29 (Newman et al. 2015). However, insights specific for energy or climate change mitigation areas are not 30 available, nor is there much information about how other innovation metrics may react to FDI. 31 Finally, technologies could be not efficient in developing countries even if they are efficient in countries 32 at the technological frontier. For instance, technologies that are highly capital intensive and labour 33 saving will be efficient only in countries where costs of capital are low and costs of labour are high. 34 Similarly, technologies which require large number of skilled labour will be more competitive in a Do Not Cite, Quote or Distribute 16-24 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 country where skilled labour is abundant (and hence cheap) than where it is scarce (Basu and Weil 2 1998; Caselli and Coleman 2006). 3 16.2.3.4 Market failures in directing technological change 4 Market forces alone cannot deliver Pareto optimal (i.e. socially efficient) due to at least two types of 5 externalities: GHG emissions that cause climate damage and knowledge spillovers that benefit firms 6 other than the inventor. Nordhaus (2011) argues that these two problems would have to be tackled 7 separately: once the favourable intellectual property right regimes (i.e. the laws or rules or regulation 8 on protection and enforcement) are in place, a price on carbon that corrects the emission externality is 9 sufficient to induce optimal level of green technological change. Acemoglu et al. (2012) demonstrates 10 that subsidizing clean technologies (and not dirty ones) is also necessary to break the lock-in of dirty 11 technological change. Recommendations for technical changes often are based on climate 12 considerations only and neglect secondary externalities and environmental costs of technology choices 13 (such as loss of biodiversity due to inappropriate scale-up of bioenergy use). The scale of adverse side 14 effects and co-benefits varies considerably between low-carbon technologies in the energy sector 15 (Luderer et al. 2019). 16 17 16.2.4 Representation of the innovation process in modelled decarbonization pathways 18 A variety of models are used to generate climate mitigation pathways, compatible with 2°C and well 19 below 2°C targets. These include Integrated Assessment Models (IAMs), energy system models, 20 computable general equilibrium models and agent based models. They range from global (Chapter 3) 21 to national models and include both top-down and bottom-up approaches (Chapter 4). Innovation in 22 energy technologies, which comprises the development and diffusion of low-, zero- and negative- 23 carbon energy options, but also investments to increase energy efficiency, is a key driver of emissions 24 reductions in model-based scenarios. 25 16.2.4.1 Technology cost development 26 Assumptions on energy technology cost developments are one of the factors that determine the speed 27 and magnitude of the deployment in climate-energy-economy models. The modelling is informed by 28 the empirical literature estimating rates of cost reductions for energy technologies. A first strand of 29 literature relies on the extrapolation of historical data, assuming that costs decrease either as a power 30 law of cumulative production, exponentially with time (Nagy et al. 2013) or as a function of technical 31 performance metrics (Koh and Magee 2008). Another approach relies on expert estimates of how future 32 costs will evolve, including expert elicitations (Verdolini et al. 2018). 33 In these models, technology costs may evolve exogenously or endogenously (Krey et al. 2019; Mercure 34 et al. 2016). In the first case, technology costs are assumed to vary over time at some predefined rate, 35 generally extrapolated from past observed patterns or based on expert estimates. This formulation of 36 cost dynamics generally underestimates future costs (Meng et al. 2021) as, among other things, it does 37 not capture any policy-induced carbon-saving technological change or any spillover arising from the 38 accumulation of national and international knowledge (Section 16.2.2 and 16.2.3) or positive 39 macroeconomics effects of a transition (Karkatsoulis et al. 2016). The influence of cost and diffusion 40 assumptions may be evaluated through sensitivity analysis. In the second case, costs are a function of 41 a choice variable within the model. For instance, technology costs decrease as a function of either 42 cumulative installed capacity (learning-by-doing) (Seebregts et al. 1998; Kypreos and Bahn 2003) or 43 R&D investments or spillovers from other sectors and countries. 44 One factor in this ‘learning-by-researching’ is applied to a wide range of energy technologies but also 45 to model improvements in the efficiency of energy use (Goulder and Schneider 1999; Popp 2004). More 46 complex formulations include two-factor learning processes (see Section 16.2.2.1) (Criqui et al. 2015; Do Not Cite, Quote or Distribute 16-25 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Emmerling et al. 2016; Paroussos et al. 2020), multi-factor learning curves (Kahouli 2011; Yu et al. 2 2011), or other drivers of cost reductions such as economies of scale and markets (Elia et al. 2021). The 3 application of two-factor learning curves to model energy technology costs is often constrained by the 4 lack of information on public and/or private energy R&D investments in many fast-developing and 5 developing countries (Verdolini et al. 2018). The approach used to model energy technology costs 6 reductions varies across technologies, even within the same model, depending on the availability of 7 data and/or the level of maturity. Less mature technologies generally depend highly on learning-by- 8 research, whereas learning-by-doing dominates in more mature technologies (Jamasb 2007). 9 In addition to learning, knowledge spillover effects are also integrated in climate-energy-economy 10 models to reflect the fact that innovation in a given country depends also on knowledge generated 11 elsewhere (Fragkiadakis et al. 2020; Emmerling et al. 2016). Models with a more detailed representation 12 of sectors (Paroussos et al. 2020) can use spillover matrices to include bilateral spillovers and compute 13 learning rates that depend on the human capital stock and the regional and/or sectoral absorption rates 14 (Fragkiadakis et al. 2020). Accounting for knowledge spillovers in the EU for PV, wind turbines, EVs, 15 biofuels, industry materials, batteries and advanced heating and cooking appliances can lead to the 16 following results in a decarbonization scenario over the period 2020–2050 as compared to the reference 17 scenario: an increase of 1.0–1.4% in GDP, 2.1–2.3% in investment, and 0.2–0.4% in employment by 18 clean energy technologies (Paroussos et al. 2017). When comparing two possible EU transition 19 strategies - being a first-mover with strong unilateral emission reduction strategy until 2030 versus 20 postponing action for the period after 2030 - endogenous technical progress in the green technologies 21 sector can alleviates most of the negative effects of pioneering low-carbon transformation associated 22 with loss of competitiveness and carbon leakage (Karkatsoulis et al. 2016). 23 16.2.4.2 Technology deployment and diffusion 24 To simulate possible paths of energy technology diffusion for different decarbonisation targets, models 25 rely on assumptions about the cost of a given technology cost relative to the costs of other technologies 26 and its ability to supply the energy demand under the relevant energy system and physical constraints. 27 These assumptions include, for example, considerations regarding renewable intermittency, inertia on 28 technology lifetime (for instance, under less stringent temperature scenarios early retirement of fossil 29 plants does not take place), distribution, capacity and market growth constraints, as well as the presence 30 of policies. These factors change the relative price of technologies. Furthermore, technological diffusion 31 in one country is also influenced by technology advancements in other regions (Kriegler et al. 2015). 32 Technology diffusion may also be strongly influenced, either positively or negatively, by a number of 33 non-cost, non-technological barriers or enablers regarding behaviours, society and institutions 34 (Knobloch and Mercure 2016)These include network or infrastructure externalities, the co-evolution of 35 technology clusters over time (“path dependence”), the risk-aversion of users, personal preferences and 36 perceptions and lack of adequate institutional framework which may negatively influence the speed of 37 (low-carbon) technological innovation and diffusion, heterogeneous agents with different preferences 38 or expectations, multi-objectives and/or competitiveness advantages and uncertainty around the 39 presence and the level of environmental policies and institutional and administrative barriers (Iyer et al. 40 2015; Baker et al. 2015; Marangoni and Tavoni 2014; van Sluisveld et al. 2020; Napp et al. 2017; 41 Biresselioglu et al. 2020). These types of barriers to technology diffusion are currently not explicitly 42 detailed in most of the climate-energy-economy models. Rather, they are accounted for in models 43 through scenario narratives, such as the ones in the Shared Socioeconomic Pathways (Riahi et al. 2017), 44 in which assumptions about technology adoption are spanned over a plausible range of values. 45 Complementary methods are increasingly used to explore their importance in future scenarios 46 (Turnheim et al. 2015; Gambhir et al. 2019; Trutnevyte et al. 2019; Doukas et al. 2018; Geels et al. 47 2016). It takes a very complex modelling framework to include all aspects affecting technology cost 48 reductions and technology diffusion, such as heterogeneous agents (Lamperti et al. 2020), regional Do Not Cite, Quote or Distribute 16-26 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 labour costs (Skelton et al. 2020), materials cost and trade and perfect foresight multi-objective 2 optimization (Aleluia Reis et al. 2021). So far, no model can account for all these interactions 3 simultaneously. 4 Another key aspect of decarbonization regards issues of acceptability and social inclusion in decision- 5 making. Participatory processes involving stakeholders can be implemented using several methods to 6 incorporate qualitative elements in model-based scenarios on future change (Doukas and Nikas 2020; 7 van Vliet et al. 2010; Nikas et al. 2017, 2018; van der Voorn et al. 2020). 8 16.2.4.3 Implications for the modelling of technical change in decarbonization pathways 9 Although the debate is still ongoing, preliminary conclusions indicate that integrated assessment 10 models tend to underestimate innovation on energy supply but overestimate the contributions by energy 11 efficiency (IPCC 2018b). Scenarios emerging from cost-optimal climate-energy-economy models are 12 too pessimistic, especially in the case of rapidly changing technologies such as wind and batteries in 13 the past decade Conversely, they tend be too optimistic regarding the timing of action, or the availability 14 of a given technology and its speed of diffusion (Shiraki and Sugiyama 2020). Furthermore, some 15 technological and economic transformations may emerge as technically feasible from IAMs, but are not 16 realistic if taking into account political economy, international politics, human behaviours, and cultural 17 factors (Bosetti 2021). 18 There is a range of projected energy technology supply costs included in the AR6 Scenario Database 19 (Box 16.1). Variations of costs over time and across scenarios are within ranges comparable to those 20 observed in recent years. Conversely, model results show that limiting warming to 2°C or 1.5°C will 21 require faster diffusion of installed capacity of low-carbon energy options and a rapid phase out of 22 fossil-based options. This points to the importance of focusing on overcoming real-life barriers to 23 technology deployment. 24 25 START BOX 16.1 HERE 26 Box 16.1 Comparing observed energy technology costs and deployment rates with projections 27 from AR6 low-carbon pathways 28 Currently observed costs and deployment for electricity supply technologies from a variety of sources 29 are compared with projections from two different sets of scenarios contained in the AR6 Scenario 30 database: 1) reference and current policies including NDCs and 2) 2°C and well-below 2°C (AR6 model 31 database). Global aggregate costs are shown for the following technologies: Coal with CCS, Gas with 32 CCS, Nuclear, Solar PV, Onshore and Offshore Wind. 33 The decrease in forecasted capital costs is not large compared to current capital costs for most 34 technologies, and does not differ much between the two scenarios (Box 16.1, Figure 1a). For Wind 35 offshore some of the models are more optimistic than the current reality (Timilsina 2020). Several 36 sources of current solar PV costs report values that are at the low end of the AR6 model scenario 37 database. By 2050, the median technology cost forecasts decrease by between 5% for nuclear and 45- 38 52% for solar (Box 16.1, Figure 1c). 39 Median values of renewables installed capacity increase with respect to 2020 capacity in “current 40 policies” scenarios (Box 16.1, Figure 1b), where energy and climate policies are implemented in line 41 with the current NDCs. More ore stringent targets (2°C) are achieved through a higher deployment of 42 renewable technologies: by 2050 solar (wind) capacity is estimated to increase by a factor of 15 (10) 43 (Box 16.1, Figure 1c). This is accompanied by an almost complete phase out of coal (-87%). The 44 percentage of median changes in installed capacity in the current policies scenarios is within comparable Do Not Cite, Quote or Distribute 16-27 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 ranges of that observed in the last decade. In the case of the 2°C and well-below 2°C scenarios, capacity 2 installed is higher for renewable technologies and nuclear, and lower for fossil-based technologies (Box 3 16.1, Figure 1c). 4 The higher deployment in 2°C scenarios cannot be explained solely as a result of technology cost 5 dynamics. In IAMs, technology deployment is also governed by system constraints that characterize 6 both scenarios, e.g. the flexibility of the energy system, the availability of storage technologies. From a 7 modelling point of view, implementing more stringent climate policies to meet the 2°C targets forces 8 models to find solutions, even if costly, to meet those intermittency and flexibility constraints and 9 temperature target constraints. 10 11 Box 16.1, Figure 1: Global Technology cost and deployment in two groups of AR6 scenarios: (1) reference 12 and current policies including NDCs and (2) 2°C and well-below 2°C. 13 Panel a) Current capital costs are sourced from Table 1, (Timilsina 2020); distribution of capital costs in 14 2030 and 2050 (AR6 database). Blue symbols represent the mean. ‘Current’ capital costs for coal and gas 15 plants with CCS are not available; Panel b) Total installed capacity in 2019 (IRENA 2020a; IEA 2020c; 16 IRENA 2020b); distribution of total installed capacity in 2030 and 2050 (AR6 database). Blue symbols 17 represent the mean; Panel c) Percentage of change in capital costs and installed capacity between (2010- 18 2020) and percentage of median change (2020-2030 and 2020-2050) (Medianyear- Do Not Cite, Quote or Distribute 16-28 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Median2020)/Median2020*100. “M” indicates the number of models, “S” the number of scenarios for which 2 this data is available. Reference and current policies' are C6 and C7 scenario categories and '2C and 3 1.5C' are C1, C2 and C3 scenario categories. Each model may have submitted data for more than one 4 model version. 5 END BOX 16.1 HERE 6 7 16.3 A systemic view of technological innovation processes 8 The innovation process, which consists of a set of sequential phases (Section 16.2.1), is often simplified 9 to a linear process. Yet, it is now well understood that it is also characterised by numerous kinds of 10 interactions and feedbacks between the domains of knowledge generation, knowledge translation and 11 application, and knowledge use (Kline and Rosenberg 1986). Furthermore, it is not just invention that 12 leads to technological change but the cumulative contribution of incremental innovations over time can 13 be very significant (Kline and Rosenberg 1986). Innovations can come not just from formal R&D but 14 also sources such as production engineers and the shop floor (Freeman 1995a; Kline and Rosenberg 15 1986). 16 This section reviews the literature focusing on innovation as a systemic process. This now predominant 17 view enriches the understanding of innovation as presented in section 16.2; it conceptualizes innovation 18 as the result of actions by, and interactions among, a large set of actors, whose activities are shaped by, 19 and shape, the context in which they operate and the user group with which they are engaging. This 20 section aligns with the discussion of socio-technical transitions (see Chapter 1 Section 1.7.3, the 21 Supplementary Material in Chapter 5, and Cross-Chapter Box 12 in this chapter). 22 23 16.3.1 Frameworks for analysing technological innovation processes 24 The resulting overarching framework that is commonly used in the innovation scholarship and even 25 policy analyses is termed as “innovation system”, where the key constituents of the systems are actors, 26 their interactions, and the institutional landscape, including formal rules, such as laws, and informal 27 restraints, such as culture and codes of conduct, that govern the behaviour of the actors (North 1991). 28 One application of this framework is that of national innovation systems (NIS), which highlights the 29 importance of national and regional relationships for determining the technological and industrial 30 capabilities and development of a country (Nelson 1993; Lundvall 1992; Freeman 1995a). Nelson 31 (1993) and Freeman (Freeman 1995a) highlight the role of institutions that determine the innovative 32 performance of national firms as way to understand differences across countries, while Lundvall (1992) 33 focuses on the “elements and relationships which interact in the production, diffusion and use of new, 34 and economically useful, knowledge”, i.e., notions of interactive learning, in which user-producer 35 relationships are particularly important (Lundvall 1988). Building on this, various other applications of 36 the “innovation systems” framework have emerged in the literature. 37 Technological Innovation systems (TIS), with technology or set of technologies (more narrowly or 38 broadly defined in different cases) as the unit of analysis and focus on explaining what accelerates or 39 hinders their development and diffusion. Carlsson and Stankiewicz (1991) define a technological 40 system as “a dynamic network of agents interacting in a specific economic/ industrial area under a 41 particular institutional infrastructure and involved in the generation, diffusion, and utilisation of 42 technology.” More recent work takes a “functional approach” to TIS (Bergek et al. 2008; Hekkert et al. 43 2007), which was later expanded with explanations of how some of the sectoral, geographical and 44 political dimensions intersect with technology innovation systems (Bergek et al. 2015; Quitzow 2015). Do Not Cite, Quote or Distribute 16-29 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Sectoral innovation systems (SIS): based on the understanding that the constellation of relevant actors 2 and institutions will vary across industrial sectors, with each sector operating under a different 3 technological regime and under different competitive or market conditions. A sectoral innovation, thus, 4 can be defined as “that system (group) of firms active in developing and making a sector's products and 5 in generating and utilising a sector's technologies” (Breschi and Malerba 1997). 6 Regional and Global innovation systems (RIS, GIS), recognising that the many innovation processes 7 have a spatial dimension, where the development of system resources such as knowledge, market 8 access, financial investment, and technology legitimacy may well draw on actors, networks, and 9 institutions within a region (Cooke et al. 1997). In other cases, the distribution of many innovation 10 processes are highly internationalised and therefore outside specific territorial boundaries (Binz and 11 Truffer 2017). Importantly, Binz and Truffer (2017) note that the GIS framework “differentiates 12 between an industry’s dominant innovation mode... and the economic system of valuation in which 13 markets for the innovation are constructed.” 14 Mission-oriented innovation systems (MIS), whose relevance comes into focus with the move towards 15 mission-oriented programs as part of the increasing innovation policy efforts to address societal 16 challenges. Accordingly, an MIS is seen as consisting of “networks of agents and sets of institutions 17 that contribute to the development and diffusion of innovative solutions with the aim to define, pursue 18 and complete a societal mission” (Hekkert et al. 2020). 19 Notably the innovation systems approach has been used in a number of climate-relevant areas such as 20 agriculture (Echeverría 1998; Klerkx et al. 2012; Horton and Mackay 2003; Brooks and Loevinsohn 21 2011), energy (Sagar and Holdren 2002; OECD 2006; Gallagher et al. 2012; Wieczorek et al. 2013; 22 Mignon and Bergek 2016; Darmani et al. 2014), industry (Koasidis et al. 2020b) and transport (Koasidis 23 et al. 2020a), and sustainable development (Clark et al. 2016; Bryden and Gezelius 2017; Anadon et al. 24 2016b; Nikas et al. 2020). 25 A number of functions can be used to understand and characterise the performance of technological 26 innovation systems (Hekkert et al. 2007; Bergek et al. 2008). The most common functions are in Table 27 16.5. 28 29 Table 16.5 Functions that the literature identified as key for well-performing technological innovation 30 systems (based on Hekkert et al (2007) and Bergek et al (2008)) Functions Description Entrepreneurial activities and Entrepreneurial activities and experimentation for translating new experimentation knowledge and/or market opportunities into real-world application Knowledge development Knowledge development includes both “learning-by-searching” and “learning-by-doing” Knowledge diffusion Knowledge diffusion through networks, both among members of a community (e.g., scientific researchers) and across communities (e.g., universities, business, policy, and users). Guidance of search Guidance of search directs the investments in innovation in consonance with signals from the market, firms or government Market formation Market formation through customers or government policy is necessary to allow new technologies to compete with incumbent technologies Resource mobilisation Resource mobilisation pertains to the basic inputs – human and financial capital – to the innovation process Do Not Cite, Quote or Distribute 16-30 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII Creation of legitimacy/counteract Creation of legitimacy or counteracting resistance to change, resistance to change through activities that allow a new technology to become accepted by users, often despite opposition by incumbent interests Development of external economies Development of external economies, or the degree to which other interests benefit from the new technology 1 2 Evidence from empirical case studies indicates that all the above functions are important and that they 3 interact with one another (Hekkert and Negro 2009). The approach therefore serves as both a rationale 4 for and a guide to innovation policy (Bergek et al. 2010). 5 A much-used, complementary systemic framework is the multilevel perspective (MLP) (Geels 2002), 6 which focuses mainly on the diffusion of technologies in relation to incumbent technologies in their 7 sector and the overall economy. A key point of MLP is that new technologies need to establish 8 themselves in a stable ‘socio-technical regime’ and are therefore generally at a disadvantage, not just 9 because their low technological maturity, but also because of an unwelcoming system. The MLP 10 highlights that the uptake of technologies in society is an evolutionary process, which can be best 11 understood as a combination of “variation, selection and retention” as well as “unfolding and 12 reconfiguration” (Geels 2002). Thus new technologies in their early stages need to be selected and 13 supported at the micro-level by niche markets, possibly through a directed process that has been termed 14 “strategic niche management” (Kemp et al. 1998). As at the macro landscape level pressures on 15 incumbent regimes mount, and those regimes destabilise, the niche technologies get a chance to get 16 established in a new socio-technical regime, which allows these technologies to grow and stabilise, 17 shaping a changed or sometimes radically renewed socio-technical regime. The MLP takes a systematic 18 and comprehensive view about how to nurture and shape technological transitions by understanding 19 them as evolutionary, multi-directional and cumulative socio-technical process playing out at multiple 20 levels over time with a concomitant expansion in the scale and scope of the transition (Elzen et al. 2004; 21 Geels 2005b). There have been numerous studies that draw on the MLP (van Bree et al. 2010; Geels et 22 al. 2017; Geels 2012) to understand different aspects of climate technology innovation and diffusion. 23 Systemic analyses of innovation have predominantly focused on industrialised countries There have 24 been some efforts to use the innovation systems lens for the developing country context (Jacobsson and 25 Bergek 2006; Lundvall et al. 2009; Altenburg 2009; Choi and Zo 2019; Tigabu 2018; Tigabu et al. 26 2015) and specific suggestions on ways for developing countries to strengthening their innovation 27 systems (e.g., by universities taking on a “developmental” role (Arocena et al. 2015) or industry 28 associations acting as intermediaries to build institutional capacities (Watkins et al. 2015; Khan et al. 29 2020), including specifically for addressing climate challenges (Sagar et al. 2009; Ockwell and Byrne 30 2016). But the conditions in developing countries are quite different, leading to suggestions that 31 different theoretical conceptualisations of the innovation systems approach may be needed for these 32 countries (Arocena and Sutz 2020), although a system perspective would still be appropriate (Boodoo 33 et al. 2018). 34 35 16.3.2 Identifying systemic failures to innovation in climate-related technologies 36 Traditional perspectives on innovation policy were mostly science-driven, and focused on strengthening 37 invention and its translation into application in a narrow sense, and a second main traditional perspective 38 on innovation policy was focused on correcting for ‘market failures’ (covered in Section 16.2) (Weber 39 and Truffer 2017). The more recent understanding of, and shift of focus to, the systemic nature on the 40 innovation and diffusion of technologies has implications for innovation policy since innovation 41 outcomes depend not just on inputs such as R&D but much more on the functioning of the overall 42 innovation system (see previous section and Section 16.4). Policies can therefore be directed at Do Not Cite, Quote or Distribute 16-31 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 innovation systems components and processes that need the greatest attention or support. This may 2 include, for example, strengthening the capabilities of weak actors and improving interactions between 3 actors (Jacobsson et al. 2017; Weber and Truffer 2017). At the same time, a systemic perspective also 4 brings into sharp relief the notion of ‘system failures’ (Weber and Truffer 2017). 5 Systemic failures include infrastructural failures; hard (e.g., laws, regulation) and soft (e.g., culture, 6 social norms) institutional failures; interaction failures (strong and weak network failures); capability 7 failures relating to firms and other actors; lock-in; and directional, reflexivity, and coordination failures 8 (Klein Woolthuis et al. 2005; Chaminade and Esquist 2010; Weber and Rohracher 2012; Wieczorek 9 and Hekkert 2012; Negro et al. 2012). By far most of the literature that unpacks such failures and 10 explores ways to overcome them is on energy-related innovation policy. For example, Table 16.6 11 summarizes a meta-study (Negro et al. 2012) that examined cases of renewable energy technologies 12 trying to disrupt incumbents across a range of countries to understand the roles, and relative importance, 13 of the ‘systemic problems’ highlighted in Section 16.3.1. 14 15 Table 16.6 Examination of systemic problems preventing renewable energy technologies from reaching 16 their potential, including number of case studies in which the particular ‘systemic problem’ was 17 identified Source: (Negro et al. 2012). Systemic problems Empirical sub-categories No. of cases Hard institutions - ‘Stop and go policy’: lack of continuity and long-term 51 regulations; inconsistent policy and existing laws and regulations -‘Attention shift’: policy makers only support technologies if they contribute to the solving of a current problem -‘Misalignment’ between policies on sector level such as agriculture, waste, and on governmental levels, i.e. EU, national, regional level, etc. -“Valley of Death”: lack of subsidies, feed-in tariffs, tax exemption, laws, emission regulations, venture capital to move technology from experimental phase towards commercialisation phase Market structures - Large-scale criteria 30 - Incremental/near-to-market innovation - Incumbent’s dominance Soft institutions - Lack of legitimacy 28 - Different actors opposing change Capabilities/capacities - Lack of technological knowledge of policy makers and engineers 19 - Lack of ability of entrepreneurs to pack together, to formulate clear message, to lobby to the government - Lack of users to formulate demand - Lack of skilled staff Knowledge infrastructure - Wrong focus or not specific courses at universities knowledge 16 institutes - Gap/Misalignment between knowledge produce at universities and what needed in practice Do Not Cite, Quote or Distribute 16-32 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII Too weak interactions - Individualistic entrepreneurs 13 - No networks, no platforms - Lack of knowledge diffusion between actors - Lack of attention for learning-by-doing Too strong interactions - Strong dependence on government action or dominant partners 8 (incumbents) - Networks allows no access to new entrants Physical infrastructure - No access to existing electricity or gas grid for RETs 2 - No decentralised, small-scale grid - No refill infrastructure for biofuels, ABG, H2, biogas 1 Depending on the sector, specific technology characteristics, and national and regional context, the 2 relevance of these systemic problems varies (Trianni et al. 2013; Bauer et al. 2017; Wesseling and Van 3 der Vooren 2017; Koasidis et al. 2020b,a), suggesting that the innovation policy mix has to be tailor- 4 made to respond to the diversity of systemic failures (Rogge et al. 2017). An illustration of how such 5 systemic failures have been addressed is given in Box 16.2, which shows how the Indian government 6 designed its standards and labelling programme for energy-efficient air conditioners and refrigerators. 7 The success of this program resulted from the careful attention to bring on board and coordinate the 8 relevant actors and resources, the design of the standards, and ensuring effective administration and 9 enforcement of the standards (Malhotra et al. 2021). 10 11 START BOX 16.2 HERE 12 Box 16.2: Standards and Labelling (S&L) for energy efficient refrigerators and air conditioners 13 in India2 14 Energy efficiency is often characterised as a “low-hanging fruit” for reducing energy use. However, 15 systemic failures such as lack of access to capital, hidden costs of implementation, and imperfect 16 information can result in low investments into adoption and innovation in energy efficiency measures 17 (Sorrell et al. 2004). To address such barriers, India’s governmental Bureau of Energy Efficiency (BEE) 18 introduced the Standards and Labelling (S&L) programme for promotion of innovation in energy 19 efficient appliances in 2006 (Sundaramoorthy and Walia 2017). While context-dependent, the 20 programme design, policies and scale-up contain lessons for addressing systemic failures elsewhere too. 21 Program design and addressal of early systemic barriers 22 To design the S&L program, BEE drew on the international experiences and technical expertise of the 23 Collaborative Labelling and Appliance Standards Program (CLASP) – a non-profit organisation that 24 provides technical and policy support to governments in implementing S&L programs. For example, 25 since there was no data on the efficiency of appliances in the Indian market, CLASP assisted with early 26 data collection efforts, resulting in a focus on refrigerators and air conditioners (ACs) (McNeil et al. 27 2008). FOOTNOTE 2 This section draws on “The role of capacity-building in policies for climate change mitigation and sustainable development: The case of energy efficiency in India” (Malhotra et al. 2021) Do Not Cite, Quote or Distribute 16-33 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Besides drawing from international knowledge, the involvement of manufacturers, testing laboratories, 2 and customers was crucial for the functioning of the innovation system. 3 To involve manufacturers, BEE employed three strategies to set the standards at an ambitious yet 4 acceptable level. First, BEE enlisted IIT Delhi (a public technical university) to engage with 5 manufacturers and to demonstrate cost-effective designs of energy-efficient appliances. Second, BEE 6 agreed to make the standards voluntary from 2006 to 2010. In return, the manufacturers agreed to 7 mandatory and progressively more stringent standards starting in 2010. Third, BEE established a multi- 8 stakeholder committee with representation from BEE, the Bureau of Indian Standards, appliance 9 manufacturers, test laboratories, independent experts, and consumer groups (Jairaj et al. 2016) to ensure 10 that adequately stringent standards are negotiated every two years. 11 At this time, India had virtually no capacity for independent testing of appliances. Here too, BEE used 12 multiple approaches towards creating the actors and resources needed for the innovation system to 13 function. First, BEE funded the Central Power Research Institute (CPRI) – a national laboratory for 14 applied research, testing and certification of electrical equipment – to set up refrigerator and AC testing 15 facilities. Second, they invited bids from private laboratories, thus creating a demand for testing 16 facilities. Third, BEE developed testing protocols in partnership with universities. Australian standards 17 for testing frost-free refrigerators were adopted until local standards were developed. Thus, once the 18 testing laboratories, protocols and benchmark prices for testing were in place, the appliance 19 manufacturers could employ their services. 20 Finally, a customer outreach program was conducted from 2006 to 2008 to inform customers regarding 21 energy efficient appliances, to enable them to interpret the labels correctly, and to understand their 22 purchase decisions and information sources (Joshi et al. 2019; Jain et al. 2018). BEE initiated a capacity 23 building program for retailers to be an information source for costumers. A comprehensive document 24 with details of different models and labels was provided to retailers, together with a condensed booklet 25 to be shared with customers. 26 Adapting policies to technologies and local context 27 While many of India’s standards and testing protocols were based on international standards, they 28 needed to be adapted to the Indian context. For example, because of higher temperatures in India, the 29 reference outside temperature of 32°C for refrigerators was changed to 36°C. 30 AC testing protocols also had to be adapted because of the emergence of inverter-based ACs. Existing 31 testing done only at a single temperature did not value inverter-based ACs’ better average performance 32 as compared to fixed-speed ACs over a range of temperatures. Thus, the Indian Seasonal Energy 33 Efficiency Ratio (ISEER) was developed for Indian temperature conditions in 2015 by studying ISO 34 standards and through consultations with manufacturers (Mukherjee et al. 2020). 35 These measures had multiple effects on technological change. As a result of stringent standards, India 36 has some of the most efficient refrigerators globally. In the case of ACs, the ISEER accelerated 37 technological change by favouring inverter-based ACs over fixed-speed ACs, driving down their costs 38 and increasing their market shares (BEE 2020). 39 Scaling up policies for market transformation 40 As the S&L program was expanded, BEE took measures to standardise, codify and automate it. For 41 example, to process a high volume of applications for labels efficiently, an online application portal 42 with objective and transparent certification criteria was created. This gave certainty to the 43 manufacturers, enabling diversity and faster diffusion of energy-efficient appliances. Thus by 2019, the 44 program expanded to cover thousands of products across 23 appliance types (BEE 2020). Do Not Cite, Quote or Distribute 16-34 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Besides issuing labels, the enforcement of standards also needed to be scaled up efficiently. BEE 2 developed protocols for randomly sampling appliances for testing. Manufacturers were given a fixed 3 period to rectify products that did not meet the standards, failing which they would be penalised and 4 the test results would be made public. 5 END BOX 16.2 HERE 6 7 16.3.3 Indicators for technological innovation 8 Assessing the state of technological innovation helps understanding how current efforts and policies are 9 doing in relation to stated objectives and how we might design policies in order to do better. 10 Traditionally, input measures such as RD&D investments and output measures such as scientific 11 publication and patents were used to characterise innovation activities (Freeman and Soete 2009), partly 12 because of the successes of specialised R&D efforts (Freeman 1995a), the predominant linear model of 13 innovation, and because such measures can (relatively) easily obtained and compared. In the realm of 14 energy-related innovation, RD&D investments remains the single most-used indicator to measure inputs 15 into the innovation process (Box 16.3). Patents counts are a widely used indicator of the outputs of the 16 innovation process, especially because they are detailed enough to provide information on specific 17 adaptation and mitigation technologies. Mitigation and adaptation technologies have their own 18 classification (Y02) with the European Patent Office (EPO) (Veefkind et al. 2012; Angelucci et al. 19 2018), which can be complemented with keyword search and manual inspection (Persoon et al. 2020; 20 Surana et al. 2020b). However, using energy-related patents as indicator of innovative activities is 21 complicated by several issues (Haščič and Migotto 2015; Jaffe and de Rassenfosse 2017; de 22 Rassenfosse et al. 2013), including the fact that the scope of what are to be considered climate mitigation 23 inventions is not always clear or straightforward. 24 Conversely, private energy R&D investments and investments by financing firms cannot be precisely 25 assessed for a number of reasons, including limited reporting and the difficulty of singling out energy- 26 related investments. This inability to precisely quantify private investments in energy R&D leads to a 27 patchy understanding of the energy innovation system, and how private energy R&D investments 28 responds to public energy R&D investments. Overall, evidence shows that some of the industrial sectors 29 that are important for meeting climate goals (electricity, agriculture and forestry, mining, oil and gas, 30 and other energy-intensive industrial sectors) are investing relatively small fractions of sales on R&D 31 (medium evidence, high agreement) (European Commission 2015; National Science Board 2018; 32 American Energy Innovation Council 2017; Jasmab and Pollitt 2005; Sanyal and Cohen 2009; Jamasb 33 and Pollitt 2008; Gaddy et al. 2017). 34 Financing firms also play an important role in the energy innovation process, but data availability is 35 also limited. The venture capital (VC) financing model, used to overcome the “valley of death” in the 36 biotech and IT space (Frank et al. 1996), has not been as suitable for hardware start-ups in the energy 37 space: for example, the percentage of exit outcomes in clean-tech start-ups was almost half of that in 38 medical start-ups and less than a third of software investments (Gaddy et al. 2017). The current VC 39 model and other private finance do not sufficiently cover the need to demonstrate energy technologies 40 at scale (Mazzucato 2013; Anadón 2012; Nemet et al. 2018). This greater difficulty in reaching the 41 market compared to other sectors may have contributed to a reduction in private equity and venture 42 capital finance for renewable energy technologies after the boom of the late 2000s (Frankfurt School- 43 UNEP Centre/BNEF 2019). 44 START BOX 16.3 HERE 45 Box 16.3 Investments in public energy R&D Do Not Cite, Quote or Distribute 16-35 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 Public energy R&D investments are a crucial driver of energy technology innovation (Sections 16.2.1.1, 2 16.4.1). Box 16.3, Figure 1 shows the time profile of energy-related RD&D budgets in OECD countries 3 as well as some key events which coincided with developments of spending (IEA 2019). Such data on 4 other countries, in particular developing countries, are not available, although recent evidence suggests 5 that expenditures are increasing there (IEA 2020c). The IEA collected partial data from China and India 6 in the context of Mission Innovation, but this is available only starting in 2014 and thus not included in 7 the Figure. 8 The figure illustrates two points. First, energy-related RD&D has risen slowly in the last 20 years, and 9 is now reaching levels comparable with the peak of energy RD&D investments following the two oil 10 crises. Second, over time there has been a reorientation of the portfolio of funded energy technologies 11 away from nuclear energy. In 2019, around 80% of all public energy RD&D spending was on low- 12 emission technologies – energy efficiency, CCUS, renewables, nuclear, hydrogen, energy storage and 13 cross-cutting issues such as smart grids. A more detailed discussion of the time profile of RD&D 14 spending in IEA countries, including as a share of GDP, are available in (IEA 2020b). Public energy RD&D investments for IEA countries 25,000 USD (2019 prices and exchange rates) 20,000 Unallocated Other cross-cutting 15,000 technologies/research Other power and storage technologies Hydrogen and fuel cells 10,000 Nuclear Renewables 5,000 Fossil fuels 0 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 15 16 Box 16.3, Figure 1 Fraction of public energy RD&D spending by technology over time for IEA (largely 17 OECD) countries between 1974 and 2018. 18 Sources: IEA RD&D Database, 2019 (IEA 2019). (extracted on November 11, 2020). 19 END BOX 16.3 HERE 20 Quantitative indicators such as energy-related RD&D spending are insufficient for the assessment of 21 innovation systems (David and Foray 1995): they only provide a partial view into innovation activities, 22 and one that is potentially misleading (Freeman and Soete 2009). Qualitative indicators measuring the 23 more intangible aspects of the innovation process and system are crucial to fully understand the 24 innovation dynamics in a climate or energy technologies or sectors (Gallagher et al. 2006), including in 25 relation to adopting an adaptive learning strategy and supporting learning through demonstration 26 projects (Chan et al. 2017). 27 In Table 16.7, both quantitative and qualitative indicators for systemic innovation are outlined, using 28 clean energy innovation as an illustrative example and drawing on a broad literature base, taking into 29 account both the input-output-outcome classification and its variations (Hu et al. 2018; Freeman and 30 Soete 1997; Sagar and Holdren 2002), combined with the functions of technological innovation systems Do Not Cite, Quote or Distribute 16-36 Total pages: 141 Final Government Distribution Chapter 16 IPCC AR6 WGIII 1 (Miremadi et al. 2018), while also being cognizant of the specific role of key actors and institutions 2 (Gallagher et al. 2012). A specific assessment of innovation may focus on part of such a list of 3 indicators, depending on what aspect of innovation is being studied, whether the analysis takes a more 4 or less systemic perspective, and the specific technology and geography considered. Similarly, 5 innovation policies may be designed to specifically boost only some of these aspects, depending 6 whether a given country/region is committed to strengthen a given technology or phase. 7 The systemic approach to innovation and transition dynamics (see also Cross-Chapter Box 12 in this 8 chapter) has advanced our understanding of the complexity of the innovation process, pointing to the 9 importance of assessing the efficiency and effectiveness in producing, diffusing and exploiting 10 knowledge (Lundvall 1992), including how the existing stock of knowledge may be recombined and 11 used for new applications (David and Foray 1995). There remains a crucial need for more relevant and 12 comprehensive approaches of assessing innovation (Freeman and Soete 2009; Dziallas and Blind 2019). 13 In the context of climate mitigation, innovation is a means to an end; therefore, there is the need to 14 consider the processes by which the output of innovation (e.g., patents) are translated into real-world 15 outcomes (e.g., deployment of low-carbon technologies) (Freeman and Soete 1997; Sagar and Holdren 16 2002). Currently, a set of quantitative metrics that, collectively, can help get a picture of innovation in 17 a particular energy technology or set of energy technologies is not available. Also the understanding of 18 how to systematically use qualitative indicators to characterise the more intangible aspects of the energy 19 innovation system and to improve front-end innovation decisions is still lacking (Dziallas and Blind 20 2019). 21 Do Not Cite, Quote or Distribute 16-37 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Table 16.7 Commonly used quantitative innovation metrics, organized by inputs, outputs and outcomes. Based on (Sagar and Holdren 2002; Gallagher et al. 2006; 2 Hekkert et al. 2007; Gallagher et al. 2012; Miremadi et al. 2018; Hu et al. 2018; Gallagher et al. 2011; Avelino et al. 2019; Gruebler et al. 2012) Structural and Function Input indicators Output indicators Outcome indicators Actors Policies systemic indicators Higher education Research programs Well-defined Scientific publications Number of technologies Governments investments and strategies processes to developed (proof-of- Private define research concept/prototypes) IPR policies R&D investments Highly-cited publications corporations priorities Knowledge development International Increase in number of Number of researchers Patents Universities technical norms Stakeholder researchers (e.g. standards) involvement in R&D projects over New product Higher education priority-setting Learning rates time configurations policies Citations to literature or Number of licensed R&D networks Development of Accessibility to patents patents Governments communication exchange Number of research Public-private co- Number of technologies centres programs agreement publications transferred Number of research Private Co-patenting Facilitation of the Strength of exchange programs Knowledge-intensive corporations development of linkage among Number of scientific Number of co-developed services exports Scientific Knowledge networks key stakeholders diffusion conferences products societies Number of patent Open-access Participation to International scientific co- Universities applications by publication policies framework publications foreigners IPR policies agreements International policy: Number of workshops Number of researchers e.g. treaties, clean ICT access and conferences working internationally development mechanism Targets set by Policy action plans Level of media coverage Budget allocations Governments government of Media strength and long-term targets industry Guidance of Shared strategies and Scenarios and foresight Mission-oriented Interest groups Innovation policies search roadmaps projects innovation programs Do Not Cite, Quote or Distribute 16-38 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII Articulation of interest Credible political from lead customers support Expectations of Media markets/profits Number of green Financial resources Access to finance projects/technologies Employment in Governments support Resource funded knowledge-intensive mobilization activities Development of Graduate in STEMS innovative financing Share of domestic credit granted to low-carbon Private firms International Gross expenditures on technology projects Employment in relevant agreements (e.g. R&D/total industries technology expenditures agreements) Share of domestic credit Private investors granted to projects Domestic credit to Scale of innovative (angel, venture Infrastructure developing private sector activities capital, private support complementary equity) assets/infrastructure Number of researchers Rate of growth of Project/program Banks in R&D per capita dedicated investment evaluation Public energy R&D Availability of expenditures/total complementary assets Innovation policies expenditures and infrastructure Expenditure on Higher education education policies Investment in complementary assets and/or infrastructure (e.g. Charging Do Not Cite, Quote or Distribute 16-39 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII infrastructure for EVs, smart grids) Venture capital on deals Ease of starting a No. of new entrants SMEs introducing Private firms business product or process Entrepreneurial % of clean energy innovation Government Risk-capital policies activities start-ups/incumbents Risk-capital Start-up support Market introduction of access to finance for providers programs new technological clean-tech start up Start-up support products Philanthropies Incubator programs services Number of new businesses Experimental application projects Creative goods exports Environmental Resource performance endowments Private firms Environmental and Market penetration of new Level of environmental Attractiveness of Governments Energy Regulation technologies impact on society renewable energy institutions Fiscal and financial infrastructure Increase in installed Renewable energy jobs regulating trade, incentives Public market support Coordination capacity finance, Market formation Renewable energy Cleantech-friendly across relevant High-tech imports investment, No of niche markets production policy processes actors (e.g., environment, Number of technologies Trade of energy renewable energy development, Transparency commercialized technology and producers, grid security, and equipment Specific tax regimes operators, and health issues distribution High-tech exports companies) Do Not Cite, Quote or Distribute 16-40 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII Youth and public Public opinion Level of discussion/debate Regulatory quality demonstration Governments among key stakeholders Policy-maker opinion Regulatory Participatory Lobbying activities (public, firms, policy- Stakeholders Creation of Executive opinion on instruments processes makers, etc.) legitimacy Regulatory acceptance regulation Citizens Political consistency and integration Greater recognition of Environmental standards Philanthropies benefits Technology support and certification 1 Do Not Cite, Quote or Distribute 16-41 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 16.3.4 Emerging policy perspectives on systemic transformations 2 Because of the multiple market, government, system, and other failures that are associated with the 3 energy system, a range of policy interventions are usually required to enable the development and 4 introduction of new technologies in the market (Twomey 2012; Jaffe et al. 2005; Bürer and 5 Wüstenhagen 2009; Veugelers 2012; Negro et al. 2012; Weber and Rohracher 2012) and used in what 6 is termed as policy mixes (Rogge and Reichardt 2016; Rogge et al. 2020; Edmondson et al. 2020, 2019) 7 . Empirical research shows that when in the energy and environment space new technologies were 8 developed and introduced in the market, it was usually at least partly as a result of a range of policies 9 that shaped the socio-technical system (Nemet 2019a; Bunn et al. 2014; Rogge and Reichardt 2016; 10 Bergek et al. 2015) (robust evidence, high agreement). An example of this systemic and dynamic nature 11 of policies is the 70-year innovation journey of solar PV, covering multiple countries, which is reviewed 12 in Box 16.4. 13 14 START BOX 16.4 HERE 15 Box 16.4 Sources of cost reductions in solar photovoltaics 16 No single country persisted in developing solar PV. Five countries each made a distinct 17 contribution. Each leader relinquished its lead. The free flow of ideas, people, machines, finance, 18 and products across countries explains the success of solar photovoltaics (PV). Barriers to 19 knowledge flow delay innovation. 20 Solar PV has attracted interest for decades, and until recently was seen as an intriguing novelty, serving 21 a niche, but widely dismissed as a serious answer to climate change and other social problems associated 22 with energy use. Since AR5, PV has become a substantial global industry—a truly disruptive 23 technology that has generated trade disputes among superpowers, threatened the solvency of large 24 energy companies, and prompted reconsideration of electric utility regulation rooted in the 1930s. More 25 favourably, its continually falling costs and rapid adoption are improving air quality and facilitating 26 climate change mitigation. PV is now so inexpensive that it is important in an expanding set of countries. 27 In 2020, 41 countries, in 6 continents, had installed at least 1GW of solar each (IRENA 2020a). 28 The cost of generating electricity from solar PV is now lower in sunny locations than running existing 29 fossil fuel power plants (Chapter 6) (IEA 2020c). Prices in 2020 were below where even the most 30 optimistic experts expected they would be in 2030. 31 The costs of solar PV modules have fallen by more than a factor of 10,000 since they were first 32 commercialised in 1957. This four orders of magnitude cost reduction from the first commercial 33 application in 1958 until 2018 can be summarised as the result of distinct contributions by the US, 34 Japan, Germany, Australia, and China—in that sequence (Green 2019; Nemet 2019c). As shown in Box 35 16.4 Figure 1, PV improved as the result of: 36 1) Scientific contributions in the 1800s and early 1900s, in Europe and the US, that provided a 37 fundamental understanding of the ways that light interacts with molecular structures, leading to the 38 development of the p-n junction to separate electrons and holes (Einstein 1905; Ohl 1941); 39 2) A breakthrough at a corporate laboratory in the US in 1954 that made a commercially available PV 40 device available and led to the first substantial orders, by the US Navy in 1957 (Gertner 2013; Ohl 41 1946); 42 3) A government R&D and public procurement effort in the 1970s in the US, that entrained skilled 43 scientists and engineers into the effort and stimulated the first commercial production lines (Laird 44 2001; Christensen 1985; Blieden 1999); Do Not Cite, Quote or Distribute 16-42 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 4) Japanese electronic conglomerates, with experience in semiconductors, serving niche markets in the 2 1980s and in 1994 launching the world’s first major rooftop subsidy program, with a declining rebate 3 schedule and demonstrating there was substantial consumer demand for PV (Kimura and Suzuki 4 2006); 5 5) Germany passing a feed-in tariff in 2000 that quadrupled the market for PV catalysing development 6 of PV-specific production equipment that automated and scaled PV manufacturing (RESA 2001; 7 Lauber and Jacobsson 2016); 8 6) Chinese entrepreneurs, almost all trained in Australia and using Australian-invented passivated 9 emitter rear cell technology, building supply chains and factories of gigawatt scale in the 2000s. 10 China became the world’s leading installer of PV from 2013 onward (Helveston and Nahm 2019; 11 Quitzow 2015). 12 7) A cohort of adopters with high willingness to pay, accessing information from neighbours, and 13 installer firms that learned from their installation experience, as well as that of their competitors to 14 lower soft costs (Gillingham et al. 2016; Ardani and Margolis 2015). 15 16 17 Box 16.4, Figure 1 Milestones in the development of low-cost solar photovoltaics (Nemet 2019c) 18 As this evolution makes clear, no individual country persisted in leading the technology and every world 19 leading firm lost its lead within a few years (Green 2019). Solar followed an overlapping but sequential 20 process of technology creation, market creation and cost reductions (comparable to emergence, early 21 adoption, diffusion and stabilisation in Cross-Chapter Box 12 in this chapter). In the technology creation 22 phase examples of central processes include flows of knowledge from one person to another, between 23 firms, and between countries as well as US and Japanese R&D funding in the 1970s and early 1980s. 24 During market creation, PVs modular scale allowed it to serve a variety of niche markets from satellites 25 in the 1950s to toys in the 1980s, when Germany transformed the industry from niche to mass market 26 with its subsidy program that began in 2000 and became important for PV in 2004. The dramatic 27 increase in size combined with its 20-year guaranteed contracts reduced risk for investors and created 28 confidence in PVs long term growth. Supportive policies also emerged outside Germany, in Spain, Italy, 29 California, and China, which spread the risk even as national policy support was more volatile. Rapid 30 and deep cost reductions were made possible by learning-by-doing in the process of operating, 31 optimising, and combining production equipment; investing and improving each manufacturing line to 32 gradually scale up to massive sizes and incremental improvements in the PV devices themselves. 33 Central to PV development has been its modularity, which provided two distinct advantages: access to 34 niche markets, and iterative improvement. Solar has been deployed as a commercial technology across 35 9 orders of magnitude: from a 1W cell in a calculator to a 1GW plant in the Egyptian desert, and almost 36 every scale in between. This modular scale enabled PV to serve a sequence of policy-independent niche 37 markets (such as satellites and telecom applications), which generally increased in size and decreased Do Not Cite, Quote or Distribute 16-43 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 in willingness to pay, in line with the technology cost reductions. This modular scale also enabled a 2 large number of iterations, such that in 2020 over three billion solar panels have been produced. 3 Compared to, for instance, approximately 1000 nuclear reactors that were ever constructed, a million 4 times more opportunities for learning-by-doing were available to solar PV: to make incremental 5 improvements, to introduce new manufacturing equipment, to optimise that equipment, and to learn 6 from failures. More generally, recent work has point to the benefits of modularity in the speed of 7 adoption (Wilson et al. 2020) and learning rates (Sweerts et al. 2020). 8 While many technologies do not fit into the solar model, some, including micro nuclear reactors and 9 direct air capture, also have modular characteristics that make them suitable for following solar’s path 10 and benefit from solar’s drivers. However, PV took solar 60 years to become cheap, which is too slow 11 for addressing climate change if a technology is now still at the lab scale. A challenge in learning from 12 the solar model is therefore to how to use public policy to speed up innovation over much shorter time 13 frames, e.g. 15 or less years. 14 END BOX 16.4 HERE 15 16 There are many definitions of policy mixes from various disciplines (Rogge et al. 2017), including 17 environmental economics (Lehmann 2012), policy studies (Kern and Howlett 2009) and innovation 18 studies. Generally speaking, a policy mix can be characterised by a combination of building blocks, 19 namely elements, processes and characteristics, which can be specified using different dimensions 20 (Rogge and Reichardt 2016). Elements include (i) the policy strategy with its objectives and principal 21 plans and (ii) the mix of policy instruments, and (iii) instrument design. The content of these elements 22 is the result of policy processes. Both elements and processes can be described by their characteristics 23 in terms of the consistency of the elements, the coherence of the processes, and the credibility and 24 comprehensiveness of the policy mix in different policy, governance, geography and temporal context 25 (Rogge and Reichardt 2016). Other aspects in the evaluation of policy mixes include framework 26 conditions, the type of policy instrument and the lower level of policy granularity, namely design 27 elements or design features (del Río 2014; del Río and Cerdá 2017). In addition, many have argued the 28 need to craft policies that affect different actors in the transition, some supporting and some 29 ‘destabilising’ (see e.g. Kivimaa and Kern (2016) and Geels (2002)). 30 Learning from the innovation systems literature, some of the recent policy focus is not only directed on 31 innovation policies that can optimise the innovation system to improve economic competitiveness and 32 growth but also policies that can induce strategic directionality and guide processes of transformative 33 changes towards desired societal objectives (Mitcham 2003; Steneck 2006). Therefore, the aim is to 34 connect innovation policy with societal challenges and transformative changes through engagement 35 with a variety of actors and ideas and incorporating equity, nowadays often referred to as a just transition 36 (Newell and Mulvaney 2013; Swilling et al. 2016; Heffron and McCauley 2018; Jasanoff 2018) (see 37 Chapter 1 and Chapter 17). This new policy paradigm is opening up a new discursive space, shapes 38 policy outcomes, and is giving rise to the emerging idea of transformative innovation policy (Diercks 39 et al. 2019; Fagerberg 2018). 40 Transformative innovation policy has a broader coverage of the innovation process with a much wider 41 participation of actors, activities and modes of innovation. It is often expressed as socio-technical 42 transitions (Elzen et al. 2004; Turnheim and Sovacool 2020) or societal transformations (Scoones 2015; 43 Roberts et al. 2018). Transformative innovation policy encompasses different ideas and concepts that 44 aim to address the societal challenges involving a variety of discussions including social innovation 45 (Mulgan 2012), complex adaptive systems (Gunderson and Holling 2002), eco-innovation (Kemp 2011) 46 and framework for responsible innovation (Stilgoe et al. 2013), value-sensitive design (Friedman and 47 Hendry 2019) and social-technical integration (Fisher et al. 2006). Do Not Cite, Quote or Distribute 16-44 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 16.4 Innovation policies and institutions 2 Building on the frameworks for identifying market failures (Section 16.2) and systemic failures (Section 3 16.3) in the innovation system for climate-related technologies, Section 16.4 proceeds as follows. First, 4 it considers some of the policy instruments introduced in Chapter 13 that are particularly relevant for 5 the pace and direction of innovation in technologies for climate change mitigation and adaptation. 6 Second, it explains why governments put in place policies to promote innovation in climate related 7 technologies. Third, it takes stock of the overall empirical and theoretical evidence regarding the 8 relationship between policy instruments with a direct and an indirect impact on innovation outcomes 9 (including intellectual property regimes) and also other outcomes (competitiveness and distributional 10 outcomes). Fourth, it assesses the evidence on the impact of trade-related policies and of sub-national 11 policies aiming to develop cleantech industrial clusters. 12 This section focuses on innovation policies and institutions which are implemented at the national level. 13 Whenever relevant, this section highlights examples of policies or initiatives that delve more deeply 14 into the main high-level sectors: power, transport, industry, buildings, and AFOLU. Whenever possible, 15 this section also discusses issues in policy selection, design, and implementation that have been 16 identified as more relevant in developing countries and emerging economies. 17 Overall, this section shows that national and subnational policies and institutions are one of the main 18 factors determining the redirection and acceleration of technological innovation and low-emission 19 technological change (Åhman et al. 2017; Rogge and Reichardt 2016; Anadon et al. 2016b; Anadón et 20 al. 2017; Roberts et al. 2018) (robust evidence, high agreement). Both technology push (e.g., scientific 21 training, R&D) and demand pull (e.g., economic and fiscal support and regulatory policy instruments), 22 as well as instruments promoting knowledge flows and especially research-firm technology transfer, 23 can be part of the mix (robust evidence, medium agreement) (see also Sections 16.2 and 16.3). 24 Public R&D investments in energy and climate-related technologies have a positive impact on 25 innovation outcomes (medium evidence, high agreement). The evidence on procurement is generally 26 positive but limited. The record economic policy instruments that can be classified as market pull 27 instruments when it comes to the competitiveness outcome (at least in the short term) is more mixed. 28 The review of the literature in this section shows that market pull policy instruments had positive but 29 also some negative impacts on outcomes in some instances on some aspects of competitiveness and 30 distributional outcomes (medium evidence, medium agreement) (Peñasco et al. 2021). For several of 31 them, carbon taxes or feed-in tariffs for example, the evidence of a positive impact on innovation is 32 more consistent than the others. Evidence suggests that complementary policies or improved policy 33 design can mitigate such short-term negative distributional impacts. 34 35 16.4.1 Overview of policy instruments for climate technology innovation 36 Government policies can influence changes in technologies, as well as changes to the systems they 37 support (Somanathan et al. 2014) (see Chapter 13 and Sections 16.2 and 16.3). 38 Technology-push policy instruments stimulate innovation by increasing the supply of new knowledge 39 through funding and performing research; increasing the supply of trained scientists and engineers 40 which contribute to knowledge-generation and provide technological opportunities, which private firms 41 can decide to commercialise (Mowery and Rosenberg 1979; Anadon and Holdren 2009; Nemet 2009b; 42 Mazzucato 2013). 43 Governments can also stimulate technological change through demand-pull (or: market-pull) 44 instruments which support market creation or expansion and technology transfer and thus promoting 45 learning by doing, economies of scale, and automation (Section 16.2). Demand-pull policy instruments Do Not Cite, Quote or Distribute 16-45 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 include regulation, carbon prices, subsidies that reduce the cost of adoption, public procurement, and 2 intellectual property regulation. Typically, technology push is especially important for early-stage 3 technologies, characterised by higher uncertainty and lower appropriability (see Section 16.2); demand- 4 pull instruments become more relevant in the later stages of the innovation process (Section 16.2) 5 (Mowery and Rosenberg 1979; Anadon and Holdren 2009; Nemet 2009b). 6 The second column of Table 16.8 summarises the set of policies shaping broader climate outcomes over 7 the past few decades in many countries outlined in Chapter 13 Section 13.6, which groups them into 8 economic and financial, regulatory, and soft instruments. Other policies, such as monetary, banking and 9 trade policies, for instance, can also shape innovation but most government action to shape energy has 10 not focussed on them. As Table 16.8 shows, this section discusses the set of policy instruments on 11 innovation outcomes, or a subset of the ‘Transformative Potential’ criterion presented in Chapter 13, 12 and thus complements the more general discussion presented there. Table 16.8 specifically prioritizes 13 the impact of the subset of policy instruments on innovation outcomes for which evidence is available. 14 This focus is complemented by a discussion of the impact of the same policy instruments on 15 competitiveness (a subcomponent of the economic effectiveness evaluation criterion) and on 16 distributional outcomes. Many of the policy instrument types listed in Table 16.8 have been 17 implemented or proposed to address different types of market or systemic failures or bottlenecks 18 described in Section 16.2 and Section 16.3 (OECD 2011a). 19 20 Table 16.8 Overview of policy instrument types covered in Chapter 13 and their correspondence to the 21 subset of policy instrument types reviewed in Chapter 16 with a focus on innovation outcomes. High- level Lower level policy Policy instrument types reviewed in Section 16.5 (for categorisation instrument type in definitions see Peñasco et al (2021)) Chapter 13 R&D investments R&D investments (including demonstration) (see Box 16.4 in Section 16.4) Subsidies for mitigation Feed in tariffs or premia (set administratively) Energy auctions Other public financing options (public investment banks, loans, loan guarantees) Economic or financial policy Emissions trading schemes Emissions trading scheme instrument types Carbon taxes Taxes/tax relief (including carbon taxes, energy taxes and congestion taxes) Government provision Government provision (focus on innovation procurement) Removing fossil fuel Not covered subsidies Border carbon adjustments Not covered Offsets Not covered Renewable obligations with tradeable green certificates Efficiency obligations with tradeable white certificates Performance standards Clean energy or renewable portfolio standards (electricity) (including with tradeable credits) Building codes (building efficiency codes) Regulatory policy Fuel efficiency standards instrument types Appliance efficiency standards Do Not Cite, Quote or Distribute 16-46 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII Technology standards Not covered Soft policy Divestment and disclosure Not covered instruments Voluntary agreements Voluntary agreements (public voluntary programs & negotiated agreements) Energy labels 1 Section 16.3 characterized technological innovation as a systemic, non-linear and dynamic process. 2 Figure 16.1 below presents a stylized (and necessarily incomplete view) connecting the innovation 3 process stages presented in Section 16.2, some of the key the mechanisms in technology innovation 4 systems, and some of the decarbonisation policy instruments that have been assessed in terms of their 5 impact on technological innovation outcomes in Section 16.4.4. As noted in the caption and discussed 6 in Section16.4.4, regulatory policy instruments also shape the early stages of technology development. 7 8 9 Figure 16.1 Technology innovation process and the (illustrative) and role of different public 10 policy instruments (on the right-hand side). Adapted from IEA (IEA 2020a). Note that, as 11 shown in section 16.4.4, demand pull instruments in the regulatory instrument category, for 12 instance, can also shape the early stages of the innovation process. Their position on the latter 13 stages is highlighted in this figure just because typically these instruments have been introduced 14 in latter stages of the development of the technology. 15 16 16.4.2 The drivers and politics of national policies for climate change mitigation and 17 adaptation 18 Governments around the world implement innovation policies in the energy and climate space with the 19 aim of simultaneously advancing environmental, industrial policy (or competitiveness), and security 20 goals (Surana and Anadon 2015; Meckling et al. 2017; Matsuo and Schmidt 2019; Anadón 2012; 21 Peñasco et al. 2021) (medium evidence, medium agreement). Co-benefits of policies shaping 22 technological innovation in climate-related technologies, including competitiveness, health, and 23 improved distributional impacts can be drivers of climate mitigation policy in the innovation sphere Do Not Cite, Quote or Distribute 16-47 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 (Deng et al. 2018a; Stokes and Warshaw 2017; Probst et al. 2020). This was the case for climate and 2 air pollution policies with local content requirements for different types of renewable energy projects, 3 for instance. in places including China (Lewis 2014; Qiu and Anadon 2012), India (Behuria 2020), 4 South Africa (Kuntze and Moerenhout 2012), and Canada (Genest 2014) (robust evidence, medium 5 agreement). 6 The emergence of industries and support groups can lead to more sustained support for innovation 7 policies (Schmid et al. 2020; Stokes and Breetz 2018; Meckling 2019; Meckling and Nahm 2019; 8 Meckling et al. 2015a; Schmidt and Sewerin 2017a). Conversely, policies shaping technology 9 innovation contribute to the creation and evolution of different stakeholder groups (robust evidence, 10 high agreement). Most of the literature on the role of the politics and interest groups has focused on 11 renewable energy technologies although there is some work on heating in buildings (e.g. Wesche et al. 12 2019). 13 As novel technologies become cost-competitive, opposition of incumbents usually grows, as well as the 14 dangers of lock-in that can be posed by the new winner. Addressing this involves adapting policy 15 (robust evidence, high agreement). 16 Three phases of politics in the development of policies to meet climate and industrial objectives can be 17 identified, at the top, the middle and the bottom of the experience curve (Breetz et al. 2018) (see also 18 Figure 16.1 above, and Geels (2002)). In the first phase of ‘niche market diffusion’, the politics of more 19 sustained support for a technology or set of technologies become possible after a group of economic 20 winners and ‘clean energy constituencies’ are created (Meckling et al. 2015a). When technologies grow 21 out of the niche (second phase), they pose a more serious competition to incumbents who may become 22 more vocal opponents of additional support for innovation in the competing technologies (Stokes 2016; 23 Geels 2014a). In a third phase, path-dependence in policymaking and lock-in in institutions need to 24 change to accommodate new infrastructure, the integration of technologies, the emergence of 25 complementary technologies and of new regulatory regimes (Aklin and Urpelainen 2013; Levin et al. 26 2012). 27 28 16.4.3 Indicators to assess the innovation, competitiveness and distributional outcomes 29 of policy instruments 30 If policy instruments are created to (at least partly) shape innovation for systemic transitions to a zero- 31 carbon future, they also need to be evaluated on their impact on the whole socio-technical system (Neij 32 and Åstrand 2006) and a wide range of goals, including distributional impacts and competitiveness and 33 jobs (Stern 2007; Peñasco et al. 2021). Given this and the current policy focus on green recovery and 34 green industrial policy, although we primarily focus on innovation outcomes, we assess also impacts 35 on competitiveness and equity. Table 16.9 lists the selected set of indicators used to assess the impact 36 of the policy instrument types covered in right hand side column in Table 16.8. The table does not 37 include technology diffusion or deployment because these are covered in the technological effectiveness 38 evaluation criterion in Chapter 13. As noted in section 16.2, it is very difficult to measure or fully 39 understand innovation with one or even several indicators. In addition, all indicators have strengths and 40 weaknesses and may be more relevant in some countries and sectors than in others. The literature 41 assessing the impact of different policy instruments on innovation often covers just one of the various 42 indicators listed in the second column of Table 16.9. 43 44 Table 16.9 Outcomes (first row) and indicators (second row) to evaluate the impact of policies shaping 45 innovation to foster carbon neutral economies. Do Not Cite, Quote or Distribute 16-48 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Sources: Innovation outcomes indicators are sourced from Del Rio and Cerdá (2014), and Peñasco et al (2021) 2 Grubb et al (2021); the indicators under the competitiveness and distributional effects criteria are sourced from 3 Peñasco et al (2021) Policy Instrument Innovation Competitiveness Distributional impacts Outcomes (part of Chapter 13 (part of Chapter 13 (defined in the same way ‘Transformative ‘Economic effectiveness’ as in Chapter13) potential’ evaluation evaluation criterion) criterion) Examples of indicators R&D investments, cost Industry creation, net job Level and incidence of used for each outcome improvements, learning creation, export of support costs, change in in the literature rates, patents, renewable energy spending on electricity publications, reductions technology equipment, as a % of total household in abatement costs, economic growth (GNP, spending, participation energy efficiency GDP), productivity, of different stakeholders, improvements, other other investments international equity (e.g., performance tCO2-eq per capita), characteristics, firms unequal access between reporting carbon saving large vs. small producers innovation or firms 4 5 16.4.4 Assessment of innovation and other impacts of innovation policy instruments 6 While it is very difficult to attribute a causal relationship between a particular policy instrument 7 implementation and different innovation indicators, given the complexity of the innovation system (see 8 Section 16.3) there is a large quantitative and qualitative literature aiming to identify such impact. 9 16.4.4.1 Assessment of the impact on innovation of technology push policy instruments: public 10 RD&D investments, other R&D incentives and public procurement 11 Economic and direct investment policy instrument types are typically associated with a direct focus on 12 technological innovation: R&D grants, R&D tax credits, prizes, national laboratories, technology 13 incubators (including support for business development, plans), novel direct funding instruments (e.g., 14 ARPA-E), and innovation procurement. 15 Public RD&D investments have been found to have a positive impact on different innovation in energy 16 and climate related technologies (robust evidence, high agreement), but the assessment relies almost 17 entirely on evidence from industrialised countries. Out of 17 publications focussing on this assessment, 18 only three found no relationship between R&D funding and innovation metrics (Peñasco et al. 2021; 19 Goldstein et al. 2020; Doblinger et al. 2019). Sixteen out of them used ex post quantitative methods and 20 one relied on theoretical ex ante assessment; only two of them included some non-industrialised 21 countries, with one being the theoretical analysis. The evidence available does not point to public R&D 22 funding for climate-related technologies crowding out private R&D (an important driver of innovation) 23 but instead crowding it in. Box 16.6 summarizes the evidence available of the impact of ARPA-E (a 24 public institution created in the United States in 2009 to allocate public R&D funding in energy) on 25 innovation and competitiveness outcomes. Another institution supporting energy R&D that is the 26 subject of much interest is the Fraunhofer Institute. 27 No evidence regarding the specific impact of R&D tax credits on climate mitigation or adaptation 28 technologies has been found, but it is worth noting that generally speaking, R&D tax credits are found 29 to incentivize innovation in firms in general, with an greater impact on small and medium firms (OECD 30 2020). This is consistent with the fact that most of the evidence on the positive impact of public R&D 31 support schemes covers small and medium firms (Howell 2017; Doblinger et al. 2019; Goldstein et al. 32 2020). Although there is a high level of agreement in the literature regarding the impact of R&D Do Not Cite, Quote or Distribute 16-49 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 investments on innovation outcomes in climate-related technologies, it is important to note that this 2 evidence comes from industrialised countries. This does not mean that public R&D investments in 3 energy have been found to have no impact on developing countries innovation or competitiveness 4 outcomes, but rather that we were not able to find such studies focussing on developing countries. 5 Overall, public procurement has high potential to incentivise innovation in climate technologies, but 6 the evidence is mixed, particularly in developing countries (limited evidence, medium agreement). 7 Public procurement accounted for 13 % of gross domestic products in OECD in 2013 and much more 8 in some emerging and developing economies (Baron 2016). Its main goal is to acquire products or 9 services to improve public services, infrastructures and facilities and, in some cases, to also incentivize 10 innovation. It is important to implement several steps in the public procurement procedure to improve 11 transparency, minimise waste, fraud and corruption of public fund. These steps range from the 12 assessment of a need, issuance of a tender to the monitoring of delivery of the good or service. Box 16.5 13 outlines a public procurement program that was implemented in The Netherlands in 2005 with a focus 14 on green technologies. In spite of the fact that green procurement policies have been implemented, the 15 literature assessing the innovation impact of public procurement programs is relatively limited and 16 suggests either a positive impact or no impact (Peñasco et al. 2021; Alvarez and Rubio 2015; Fernández- 17 Sastre and Montalvo-Quizhpi 2019; Baron 2016). The majority of cases where the impact is positive 18 are analyses of industrialised countries, while no impact emerges in the case of a developing country 19 (Ecuador). More empirical research is needed to understand the impact of public procurement, which 20 has the potential to support the achievement of other societal challenges (Edler and Georghiou 2007; 21 Henderson and Newell 2011; ICLEI 2018; Baron 2016) in both developing and developed countries. 22 23 START BOX 16.5 HERE 24 Box 16.5 Green Public Procurement in The Netherlands 25 In 2005, the Dutch national government acknowledged a move in the House of Representatives to utilise 26 their annual spending power to promote the market for sustainable goods and services as well as to play 27 as a role model. Hence, a policy for environmentally friendly procurement was developed and 28 implemented across the national, local and provincial governments. Subsequently, sustainable public 29 procurement has expanded into a multidimensional policy in the Netherlands, accommodating policies 30 on green public procurement, bio-based public procurement, international social criteria, social return 31 on investment, innovation-oriented public procurement and circular economy. 32 The Green Public Procurement (GPP) policy is targeted at minimising the negative impacts of 33 production and consumption on the nature environment (Melissen and Reinders 2012; Cerutti et al. 34 2016). It includes a wide range of environmental criteria for different product groups that public 35 organisations frequently procure such as office equipment, uniforms, road works and catering. There 36 are 45 product groups (Melissen and Reinders, 2012) and 6 product clusters part of the government’s 37 purchasing in terms of sustainability(PIANOo Expertisecentrum 2020). The six product clusters are: i) 38 Automation & telecommunications, ii) Energy, iii) Ground, road & hydraulic engineering, iv) Office 39 facilities and services, v) Office buildings, and vi) Transport (PIANOo Expertisecentrum 2020).The 40 GPP 2020 Tender Implementation Plan spells out the terms and conditions for making their green public 41 procurement. Some of these are confidential documents and are not shared online. Others are available 42 for download. The tender implementation plan for the Netherlands is available on 43 https://gpp2020.eu/low-carbon-tenders/open-tenders/. One of the important scenarios is that the public 44 procurers need the details of LCA analysis carried out in a tool called DuboCalc which calculates the 45 environmental impacts of the materials and methods of an infrastructural projects. GPP 2020 has 46 reported that three million tonnes of CO2 would be saved in the Netherlands alone if all Dutch public 47 authorities applied the national Sustainable Public Procurement Criteria. Do Not Cite, Quote or Distribute 16-50 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Research has been carried out to determine the prime mover for implementing Green Public 2 Procurement. An online survey of was administered among public procurement officers that subscribed 3 to the newsletters of two Dutch associations that provide advice and training to public procurers, 4 yielding a sample size of over 200 (Grandia and Voncken 2019). The first association is called NEVI 5 which is the only organisation in the Netherlands that offers certified procurement training programmes. 6 The second association is called PIANOo which is a public procurement expertise centre paid by the 7 Dutch national government for bringing together relevant information regarding public procurement 8 and providing public procurers with useful tools through their websites, workshops, meetings and 9 annual conferences. The data from the survey was then analysed using structural equations modelling 10 (SEM) and the results show that ability, motivation and opportunities affect the implementation of GPP. 11 Particularly, opportunity was found to affect green public procurement, innovation-oriented public 12 procurement and circular economy but not the other types of public procurement. 13 END BOX 16.5 HERE 14 15 16.4.4.2 Assessment of the impact on competitiveness of technology push policy instruments: public 16 RD&D investments, other R&D incentives and public procurement 17 Public R&D investments in the energy, renewables, environment space are generally associated with 18 positive impacts on industrial development or ‘competitiveness outcome’ (robust evidence, medium 19 agreement). In a number of cases negligible or negative impacts emerge (Peñasco et al. 2021; Goldstein 20 et al. 2020; Doblinger et al. 2019). The majority of these 15 analyses rely on ex post quantitative 21 methods, while only four use ex ante modelling approaches. Also, in this case, the vast majority of the 22 evidence is from industrialised countries. 23 There is limited and mixed evidence regarding the (positive or negative) impact of public procurement 24 for low-carbon or climate technologies and it emerges from developed countries (limited evidence, low 25 agreement). All of the four evaluations identified in the Peñasco et al (2021) review relied on qualitative 26 methods. One found a positive impact, another a negative impact and two others found no impact. All 27 of the studies covered European country experiences. 28 R&D and procurement policies have a positive impact on distributional outcomes (limited evidence, 29 high agreement). Peñasco et al (2021) identify three evaluations of the impact of RD&D funding on 30 distributional outcomes (two using quantitative methods and one ex ante theoretical methods) and one 31 of procurement on distributional outcomes (relying on qualitative analysis). 32 16.4.4.3 Emerging insights on different public R&D and demonstration funding schemes 33 The ability of a given R&D policy instrument to impact innovation and competitiveness depends to 34 some extent on policy design features (limited evidence, high agreement). As discussed in section 35 16.4.4.4, this is not unique to R&D funding. Most of these assessments use a limited number of 36 indicators (e.g., patents and publications and follow-on private financing, firm growth and survival, 37 respectively), focusses on the energy sector and on the US and other industrialised countries. 38 Extrapolating to emerging economies and low-income countries is difficult. There is no evidence on 39 the impact of different ways of allocating public energy R&D investments in the context of developing 40 countries. 41 Block funding, which tends to be more flexible, can lead to research that is more productive or novel, 42 but there are other factors that can affect the extent to which block funding can lead to more or less 43 novel outcomes (limited evidence, medium agreement). Research on national research laboratories, 44 which conduct at least 30% of all research in 68 countries around the world (Anadon et al. 2016a), are 45 a widespread mechanism to carry out public R&D and allocate funds, but assessments of their 46 performance is limited to developed countries. R&D priorities are also guided by institutions, and Do Not Cite, Quote or Distribute 16-51 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 research focussed on general technology innovation policy finds that institutions often do not embody 2 the goals of the poor or marginalized (Anadon et al. 2016b). 3 In the case of the US Department of Energy, block funding that can be quickly allocated to novel l 4 projects (such as that allocated to National Labs as part of the Laboratory Directed Research and 5 Development—LDRD—funding) has been found to be associated with improved innovation indicators 6 (Anadon et al. 2016a). Research on Japan on R&D funding in general (not for climate-related 7 technologies) however, indicates that R&D funds allocated competitively result in higher novelty for 8 ‘high status’ (the term used in the paper to refer to senior male researchers), while block funding was 9 associated with research of higher novelty for lower status researchers (e.g., junior female researchers) 10 (Wang et al. 2018). 11 12 START BOX 16.6 HERE 13 Box 16.6 ARPA-E a novel R&D funding allocation mechanism focussed on an energy mission 14 One approach for allocating public R&D funds in energy involves relying on active program managers 15 and having clear technology development missions that focus on high-risk high-reward areas and 16 projects. This approach can be exemplified by a relatively new energy R&D funding agency in the US, 17 the Advanced Research Projects Agency for Energy (ARPA-E). This agency was created in 2009 and 18 it was modelled on the experience of DARPA (a US government agency funding high risk high reward 19 research in defence-related areas (Bonvillian and Van Atta 2011; Bonvillian 2018; U.S. National 20 Academies of Sciences Engineering and Medicine 2017). DARPA program managers had a lot of 21 discretion for making decisions about funding projects, but since energy R&D funding is usually more 22 politically vulnerable than defence R&D funding, the ARPA-E novel involved program managers 23 requesting external review as an informational input (Azoulay et al. 2019). 24 Like DARPA, ARPA-E program managers use an ‘active management approach that involves 25 empowering program manages to make decisions about funding allocation, milestones and goals. 26 ARPA-E managers also differ from other R&D allocation mechanisms in that ARPA-E staff retained 27 some control on the funded projects after the allocation of the funds. As argued by (Azoulay et al. 2019), 28 even though this relative control over the project can result in a reduction in the flexibility of funded 29 researchers, some ‘exploration’ happens at the program manager level. 30 Research on ARPA-E also sheds light on how program managers make decisions about what projects 31 to fund: about the process of project selection. Program managers do not just follow the rankings of 32 peer reviewers (sometimes projects with very disparate rankings were funded) and in many cases 33 program managers reported using information from review comments instead of the rankings (Goldstein 34 and Kearney 2020). Azoulay et al. (2019) suggest that if expert disagreement is a useful proxy for 35 uncertainty in research, then the use of individual discretion in ARPA-E would result in a portfolio of 36 projects with a higher level of uncertainty, as defined by disagreement among reviewers. Moreover, 37 under the premise that uncertainty is a corollary to novelty, individual discretion is an antidote to novelty 38 bias in peer review. 39 While innovation is notoriously hard to track and, particularly for emerging technologies, it can take a 40 lot time to assess, early analysis has shown that this mission-orientation and more ‘actively managed’ 41 R&D funding program may yield greater innovation patenting outcomes than other US energy R&D 42 funding programs and a greater or similar rate of academic publications when compared to other public 43 funding agencies in energy in the US, ranging from the Office of Science, the more applied Office of 44 Energy Efficiency and Renewable Energy or the small grants office (Goldstein and Narayanamurti 45 2018; U.S. National Academies of Sciences Engineering and Medicine 2017). In addition research 46 analysing the first cohort of cleantech start-ups has found that start-ups supported by ARPA-E had more Do Not Cite, Quote or Distribute 16-52 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 innovative outcomes when compared to those that had applied but not received funding, with others 2 that had not received any government support, and with others that had received other types of 3 government R&D support (Goldstein et al. 2020). Overall, the mission-oriented ARPA approach has 4 been successful in the United States when it comes to innovation outcomes. The extent to which it can 5 yield the same outcomes in other geographies with different innovation and financing environments 6 remains unknown. (limited evidence, high agreement). 7 END BOX 16.6 HERE 8 9 Public financing for R&D and research collaboration in the energy sector is important for small firms, 10 at least in industrialised countries, and it does not seem to crowd out private investment in R&D 11 (medium evidence, high agreement). Small US and UK firms accrue more patents and financing when 12 provided with cash incentives for R&D in the form of grants (Pless 2019; Howell 2017). US cleantech 13 start-ups which partner with government partners for joint technology development or licensing 14 partnerships accrue more patents and follow on financing (Doblinger et al. 2019). 15 Overall, the body of literature on public R&D funding design in energy and climate related technologies 16 provides some high-level guidance on how to make the most of these direct RD&D investments in 17 energy technologies in the climate change mitigation space, including: giving researchers and technical 18 experts autonomy and influence over funding decisions; incorporating technology transfer in research 19 organisations; focussing demonstration projects on learning; incentivising international collaboration 20 in energy research; adopting an adaptive learning strategy; and making funding stable and predictable 21 (Narayanamurti and Odumosu 2016; Narayanamurti et al. 2009; Chan et al. 2017) (medium evidence, 22 high agreement). 23 Without carefully designed public funding for demonstration efforts, often in a cost shared manner with 24 industry, the experimentation at larger scales needed for more novel technologies needed for climate 25 change mitigation may not take place. (medium evidence, high agreement). Government funding 26 specifically for technology demonstration projects, for RD&D (research, development and 27 demonstration) in energy technologies plays a crucial supporting role (Section 16.2.1). Governments 28 can facilitate knowledge spill-overs between firms, between countries, and between technologies (see 29 Section 16.2, Cohen et al (2002) and Baudry and Bonnet (2019)). 30 16.4.4.4 Assessment of the impact on innovation and on competitiveness and distributional 31 outcomes of market pull policy instruments 32 Demand pull policies such as tradeable green certificates, taxes, or auctions, are essential to support 33 scaling up efforts (Remer and Mattos 2003; Nahm and Steinfeld 2014; Wilson 2012). Just like for R&D 34 investments, research has indicated that effective demand pull needs to be credible, durable, and aligned 35 with other policies (Nemet et al. 2017) and that the effectiveness of different demand pull instruments 36 depends on policy design (del Río and Kiefer 2021). Historical analyses of the relative importance of 37 demand pull and technology push are clear; both are needed to provide robust incentives for investment 38 in innovation. Interactions between them are central as their combination enables innovators to connect 39 a technical opportunity with a market opportunity (Grubler and Wilson 2013; Freeman 1995b; 40 Jacobsson et al. 2004). It is important to note that these market pull policies are often put in place 41 primarily to meet security and/or environmental goals, although innovation and competitiveness are 42 sometimes also pursued explicitly. 43 44 Emission Trading Schemes 45 Overall evidence suggests that the emissions trading schemes, as currently designed, have not 46 significantly contributed to innovation outcomes (medium evidence, medium/high agreement). Do Not Cite, Quote or Distribute 16-53 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Penasco et al. (2021) review 20 evaluations: eight identified a positive impact (although in at least two 2 cases the paper indicated the impact was small or negligible), 11 no impact and one was associated with 3 a negative impact on innovation indicators. The studies that found no impact and the studies that found 4 some impact covered all three methods covered (quantitative ex post, qualitative and theoretical and ex 5 ante analysis). Another review focussed only on empirical studies (mainly quantitative but also 6 qualitative), covered a slightly longer period and identified 19 studies (15 using quantitative methods) 7 (Lilliestam et al. 2021). With a narrower set of indicators of innovation, they concluded there was very 8 little empirical evidence linking the emissions trading schemes studied to date and innovation 9 (Lilliestam et al. 2021). This review focussed mainly on papers evaluating the earlier stages of the 10 European Emissions Trading Scheme, which featured relatively low CO2 prices and covered a small set 11 of firms, showing that carbon pricing policy design is an important determinant of innovation outcomes. 12 Combining both reviews, there are a total of 27 individual studies, some of them providing mixed 13 evidence of impact, and 23 of them suggest there was no impact or (in a couple of cases) it was small. 14 It is important to note that some researchers note that, for particular subsectors and actors, emissions 15 trading schemes have had an impact on patenting trends (Calel and Dechezleprêtre 2016). Overall the 16 expectation is that higher prices and coverage would result in higher impacts and that, over time, the 17 impact on innovation would grow. 18 Carbon and environmental taxes 19 The impact of carbon taxes on innovation outcomes is more positive than that for ETS schemes but the 20 evidence is more limited (limited evidence, medium agreement). Assessments of their impact on 21 innovation metrics have been very limited, with only four studies (three quantitative and one ex ante). 22 Three of the studies found a positive impact of carbon taxes on innovation outcomes and one found no 23 impact (Peñasco et al. 2021). 24 Depending on the design (including the value and coverage of the tax), carbon taxes can either have 25 positive, negative or null impact on competitiveness and distributional outcomes (medium evidence, 26 medium agreement). The evidence on the impact of carbon taxes on competitiveness is significant (a 27 total of 27 evaluations) and mixed, with six of them reporting some positive impacts, ten reporting no 28 impact, and 11 reporting negative impacts (so 59% were not associated with negative impacts). Most 29 of the evaluations reporting negative impacts were theoretical assessments, and only three ex post 30 quantitative analysis (Peñasco et al. 2021). 24 evaluations covered distributional impacts of carbon 31 taxes and other environmental taxes and the majority (15) found the existence of some negative 32 distributional impacts, six found positive impacts and three no distributional impacts. Differences in the 33 result of the assessments stem from the design of the taxes (Peñasco et al. 2021). It is important to note 34 that, once again, the evidence comes from industrialized countries and emerging economies. 35 Feed-in-Tariffs 36 Many factors affect the impacts of feed in tariffs on outcomes other than innovation (robust evidence, 37 high agreement). While FITs have been generally associated with positive innovation outcomes, some 38 of the differences found in the literature may arise from differences in the evaluation method (Peñasco 39 et al. 2021) or differences in policy design (e.g., the level and the rate of decrease of the tariff) 40 (Hoppmann et al. 2014), the policy mixes (Rogge et al. 2017), the technologies targeted and their stage 41 of development (Huenteler et al. 2016b), and the geographical and temporal context of where the policy 42 was put in place (Section 16.3). Research has also found that, particularly for less mature technologies, 43 a higher technology specificity in the design of FITs is associated with more innovation (Del Río 2012). 44 Feed-in-tariffs yield better results if they account for the specificities of the country; else, the technology 45 and the policy could result in negative distributional and (to a lesser extent) competitiveness impacts. 46 Meckling et al (2017) indicate that an ‘enduring challenge’ of technology-specific industrial policy such 47 as some feed-in-tariffs is to avoid locking in suboptimal clean technologies—a challenge which, among 48 other options, could be overcome with targeted niche procurement for next generation technologies. Do Not Cite, Quote or Distribute 16-54 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Other authors have cautioned that the move from renewable feed-in-tariffs to auctions may favour 2 existing PV (e.g., polysilicon) over more novel solar power technologies (Sivaram 2018b) such as thin- 3 film PV, amorphous PV, and perovskites. 4 Policy design, policy mixes, and domestic capacity and infrastructure are important factors determining 5 the extent to which economic policy instruments in industrialised countries and emerging economies 6 can also lead to positive (or at least not negative) competitiveness outcomes and distributional outcomes 7 (medium evidence, medium agreement) (Section 16.3). Prioritising low cost energy generation in the 8 design of FIT schemes can result in a lower focus of innovation efforts on more novel technologies and 9 greater barriers to incumbents in less mature technologies (Hoppmann et al. 2013). Similarly, case study 10 research from Mexico and South Africa indicates that focusing on low-cost renewable energy 11 generation only can result in a greater reliance on existing foreign value chains and capital, and thus in 12 lower or negative impacts on domestic competitiveness—in other words, some approaches can hinder 13 the development of the local capabilities that could result in greater long-term benefits domestically 14 (Matsuo and Schmidt 2019). Evidence for developing countries indicates that local and absorptive 15 capacity also play an important role in particular on the ability of policies to contribute to 16 competitiveness or industrial policy goals (e.g., Binz and Anadon 2018). Research comparing China’s 17 and India’s policies and outcomes on wind also suggest that policy durability and systemic approaches 18 can affect industrial outcomes (Surana and Anadon 2015). 19 Energy auctions 20 The evidence of the impact of renewable energy auctions on innovation outcomes is very small and 21 provides mixed results (limited evidence, low agreement). Out of six evaluations, three of them identify 22 positive impacts, two no impacts, and one negative impacts. All of the evaluations but one were 23 qualitative or theoretical and the quantitative assessment indicated no impact (Peñasco et al. 2021). 24 There is more evidence covering emerging economies analysing the impacts of auctions when 25 compared to other policy instrument types. For example, there is work comparing the approaches to 26 renewable energy auctions in South Africa and Denmark (Toke 2015) finding a positive impact on the 27 latter stages of innovation (mainly deployment), and broader work on auctions covering OECD 28 countries as well as Brazil, South Africa and China not finding a significant impact on innovation 29 (Wigand et al. 2016), and work comparing renewable energy auctions in different countries in South 30 America finding generally a positive impact on innovation outcomes (Mastropietro et al. 2014). The 31 body of evidence on the impact of auctions on competitiveness is also limited (six evaluations) and 32 indicates negative outcomes of renewable auctions of competitiveness (limited evidence, low 33 agreement). As with other policies, the design of the auctions can affect innovation outcomes (del Río 34 and Kiefer 2021). Only two studies investigated distributional outcomes and both were negative. 35 36 Other financial instruments 37 There is no explicit literature on the ability of green public banks, and targeted loans, and loan 38 guarantees to lead to upstream innovation investments and activities, although there is evidence on their 39 role in deployment (see e.g. Geddes et al (2018)). This notwithstanding the key role of these institutions 40 in the innovation system (Sections 16.2.1 and 16.3) (OECD 2015c; Geddes et al. 2018) and the belief 41 that they can de-risk scale-up and the testing of business models (Probst et al. 2021; Geddes et al. 2018) 42 (see Chapter 17). 43 44 Renewable obligations with tradeable green certificates 45 There is mixed evidence of the impact of tradeable green certificates (TGCs) on innovation (limited 46 evidence, low agreement) and competitiveness (limited evidence, low agreement). Out of the 11 47 evaluations in Peñasco et al (2021), six found no impact, two a positive impact, and three a negative 48 impact. All of them used a qualitative research approach. Of the six studies focusing on competitiveness 49 outcomes, three conclude TGCs have had no impact on competitiveness, while two indicate negative Do Not Cite, Quote or Distribute 16-55 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 impact and one a positive impact. Only one of the studies was quantitative and did not identify an impact 2 on competitiveness. 3 TGC are associated with the existence of negative distributional impacts in most applications (medium 4 evidence, high agreement). Ten out of 12 studies identify the existence of some negative impacts. All 5 but one of these studies (which focussed on India) are based on analysis of policies implemented in 6 industrialised countries. 7 Clean Energy and renewable Portfolio Standards 8 The impact of renewable portfolio standards without tradeable credits on innovation outcomes is 9 negligible or very small (medium evidence, medium agreement). Out of the nine studies, seven reported 10 no impact on innovation outcomes and two a positive impact (Peñasco et al. 2021). Most of these papers 11 focussed on patenting and private R&D innovation indicators and not cost reductions. Impact on 12 competitiveness is found to be negligible or positive (limited evidence, medium agreement). Out of 13 eight evaluations, five report positive impact and three negligible impact; only two are quantitative 14 studies (Peñasco et al. 2021). Negative distributional impacts from renewable portfolio standards can 15 emerge in some cases (limited evidence, low agreement). Out of eight evaluations, four identified 16 positive impacts, and four negative impacts; all of the studies identifying a positive impact were 17 theoretical. There are efforts focussed on clean energy portfolio standards which include technologies 18 beyond renewables. 19 20 Efficiency obligations with tradeable credits 21 The impact of tradeable white certificates in innovation is largely positive, but the evidence is limited 22 (limited evidence, medium/high agreement). Out of four evaluations, only one of which was 23 quantitative, three report positive impact and one no impact (Peñasco et al. 2021). The impact of white 24 certificates on competitiveness is positive (limited evidence, high agreement) while that on 25 distributional outcomes is very mixed (limited evidence, low agreement). Two theoretical studies report 26 positive competitiveness impacts. Out of 11 evaluations of distributional outcomes, eight rely on 27 theoretical ex ante approaches. Seven evaluations reported positive impacts (four of them using 28 theoretical methods), three of them (using theoretical methods) indicated negative impacts and one of 29 them no impact. 30 31 Building codes 32 There is evidence of the impact of building codes on innovation outcomes (Peñasco et al. 2021). Only 33 two studies assessed competitiveness impacts (one identifying positive impacts and one negligible ones) 34 and three studies identifying distributional impacts, all positive. 35 Overall, the evidence on the impact of the market pull policy instruments covered in Section 16.4.4.4 36 when it comes to the competitiveness outcome (at least in the short term) is more mixed. For some of 37 them, the evidence of a positive impact on innovation is more consistent than the others (for carbon 38 taxes or FITs, for example). Peñasco et al (2021) found that the disagreements in the evidence regarding 39 the positive, negative or no impact of a policy on competitiveness or distributional outcomes can often 40 be explained by differences in policy design, differences in geographical or temporal context (since the 41 review included evidence from countries from all over the world), or on how policy mixes may have 42 affected the ability of the research design of the underlying papers to separate the impact of the policy 43 under consideration from the others. 44 45 16.4.4.5 Assessment of the impact on innovation, competitiveness and distributional outcomes of 46 regulatory policy instruments targeting efficiency improvements 47 There is medium evidence that the introduction of flexible, performance-based environmental 48 regulation on energy efficiency in general (e.g., efficiency standards) can stimulate innovative Do Not Cite, Quote or Distribute 16-56 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 responses in firms (Ambec et al. 2013; Popp 2019) (medium evidence, high agreement). Evidence 2 comes from both observational studies that examine patenting, R&D or technological responses to 3 regulatory interventions, and from surveys and qualitative case studies in which firms report regulatory 4 compliance as a driving force for the introduction of environmentally-beneficial innovations (Grubb et 5 al. 2021). While the literature examining the impact of environmental regulation on innovation is large, 6 there have been fewer studies on the innovation effects of minimum energy or emissions performance 7 regulations specifically relating to climate mitigation. We discuss in turn two types of efficiency 8 regulations: on vehicles, and on appliances. 9 Relationship between automotive efficiency regulations and innovation 10 The announcement, introduction and tightening of vehicle fleet efficiency or GHG emission standards 11 either at the national or sub-national level positively impacts innovation as measured by patents 12 (Barbieri 2015) or vehicle characteristics (Knittel 2011; Kiso 2019) as summarised in a review by 13 Grubb et al (2021). Detailed studies on the innovation effects of national pollutant (rather than energy) 14 regulations on automotive innovation also indicate that introducing or tightening performance standards 15 has driven technological change (Lee et al. 2010). Some studies in the US that examine periods in which 16 little regulatory change took place have found that the effects of performance standards on fuel economy 17 have been small (Knittel 2011) or not significant relative to the innovation effects of prices (Crabb and 18 Johnson 2010). This is at least in part because ongoing efficiency improvements during this period were 19 offset by increases in other product attributes. For example, Knittel (2011) study observed that size and 20 power increased without a corresponding increase in fuel consumption. It has also been observed that 21 regulatory design may introduce distortions that affect automotive innovation choices: in particular, 22 fuel economy standards based on weight classes have been observed to distort light-weighting strategies 23 for fuel efficiency in both China (Hao et al. 2016) and Japan (Ito and Sallee 2018). 24 A number of studies have focused on the impacts of a sub-national technology-forcing policy: the 25 California Zero Emission Vehicle (ZEV) mandate. When it was introduced in 1990, this policy required 26 automotive firms to ensure that 2% of the vehicles they sold in 1998 would be zero emissions. In the 27 years immediately after introduction of the policy, automotive firms reported that it was a significant 28 stimulus to their R&D activity in electric vehicles (Brown et al. 1995). Quantitative evidence examining 29 patents and prototypes has indicated that the stringency of the policy was a significant factor in 30 stimulating innovation, though this was in part dependent on firm strategy (Sierzchula and Nemet 2015). 31 Like in the previous instruments, most of the evidence comes from industrialised countries and 32 additional research on other countries would be beneficial. 33 Relationship between appliance efficiency standards and innovation 34 Regulation-driven deployment of existing technologies can generate innovation in those technologies, 35 through learning-by-doing, induced R&D and other mechanisms, although not in all cases (Grubb et al. 36 2021) (medium evidence, medium agreement). The introduction or tightening of minimum energy 37 performance standards for appliances (and in the case of Noailly (2012) for buildings) have driven 38 innovation responses, using direct measures of product attributes (Newell et al. 1999) and patents 39 (Noailly 2012; Kim and Brown 2019), though not all studies have found a significant relationship 40 (Girod et al. 2017). There is also evidence of a correlation between regulation-driven deployment of 41 energy-efficient products with accelerated learning in those technologies (Van Buskirk et al. 2014; Wei 42 et al. 2017). 43 In addition to observational studies, evidence on the relationship between innovation and regulation 44 comes from surveys in which survey respondents are asked whether they have engaged in innovation 45 leading to energy saving or reduced GHG emissions, and what the motivations were for such innovation. 46 Survey evidence has found that expected or current regulation can drive both R&D investment and 47 decisions to adopt or introduce innovations that reduce energy consumption or CO2 emissions (Horbach 48 et al. 2012; Grubb et al. 2021). Survey-based studies, however, tend not to specify the type of regulation. Do Not Cite, Quote or Distribute 16-57 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Competitiveness and distributional impacts associated with vehicles and appliance performance 2 standards 3 Minimum energy performance standards and appliance standards have been known to result in negative 4 distributional impacts (limited evidence, medium/high agreement). Several studies focused on the US 5 have highlighted that minimum energy performance standards for vehicles tend to be regressive, with 6 poorer households disproportionately affected (Levinson 2019; Jacobsen 2013), particularly when 7 second-hand vehicles are taken into account (Davis and Knittel 2019). Similar arguments, though with 8 less evidence, have been made for appliance standards (Sutherland 2006). 9 Overall, the extent to which regulations in energy efficiency result in positive or negative 10 competitiveness impacts in firms is mixed (limited evidence, high disagreement). A meta-analysis of 11 107 studies, of which 13 focused on regulations relating to energy consumption or GHG emissions, 12 found that around half showed that regulations resulted in competitiveness impacts, while half did not 13 (Cohen and Tubb 2018). Cohen and Tubb (2018) also found that studies examining performance-based 14 regulations were less likely to find positive competitiveness impacts than those that examined market- 15 based instruments. 16 Insights into causal mechanisms and co-evolutionary dynamics from case studies on efficiency 17 regulations 18 While most of the literature addresses the extent to which regulation can induce innovation, a number 19 of case studies highlight that innovation can also influence regulation, as the costs of imposing 20 regulation are reduced and political interests emerge that seek to exploit competitive advantages 21 conferred by successfully developing energy-efficient or low-carbon technologies (medium evidence, 22 high agreement). Case studies map the causal mechanisms relating regulations and innovation 23 responses in specific firms or industries (Ruby 2015; Wesseling et al. 2015; Kemp 2005; Gann et al. 24 1998). 25 26 16.4.4.6 Assessment of the impact on innovation and on competitiveness and distributional 27 outcomes of soft instruments 28 Energy labels and innovation 29 The literature specifically focusing on the impacts of labels is very limited and indicates positive 30 outcomes (limited evidence, high agreement). Energy labels may accompany a minimum energy 31 performance standard and the outcomes of these policies are often combined in literature (IEA 2015). 32 But again, given the limited evidence more research is needed. Although there are many studies on 33 energy efficiency more broadly and for both standards and labels, only eight studies specifically focus 34 on labels. Furthermore, seven of them report positive outcomes and one negative outcomes. Six of the 35 studies used qualitative methods mentioning the impacts of labelling on the development of new 36 products (Wiel et al. 2006). Research specifically comparing voluntary labels with other mechanisms 37 found a significant and positive relationship between labels and the number of energy-efficient 38 inventions (Girod et al. 2017). More research is needed especially in developing countries that have 39 extensive labelling programs in place, and also with quantitative methods, to develop evidence on the 40 impacts of labelling on innovation. Box 16.7 discusses an example of a combination of policy 41 instruments in China including labelling, bans and financial support. 42 43 START BOX 16.7 HERE 44 Box 16.7 China Energy Labelling Policies, combined with sale bans and financial subsidies 45 From 1970 to 2001, China was able to significantly limit energy demand growth through energy- 46 efficiency programs. Energy use per unit of gross domestic product (GDP) declined by approximately 47 5% yr-1 during this period. However, between 2002 and 2005, energy demand per unit of GDP increased Do Not Cite, Quote or Distribute 16-58 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 on average by 3.8% yr-1. To curb this energy growth, in 2005, the Chinese government announced a 2 mandatory goal of 20% reduction of energy intensity between 2006 and 2010 (Zhou et al. 2010; Lo 3 2014). 4 An Energy Labelling System was passed in 2004. It requires the manufacturers to provide information 5 about the efficiency of their electrical appliances to consumers. From 2004 to 2010, 23 electrical 6 appliances (including refrigerators, air conditioners and flat-screen TVs) being labelled as energy 7 efficient with 5 different grades, with Grade 1 being the most energy efficient and grade 5 the least 8 efficient. Any appliances with an efficiency grade higher than 5 cannot be sold in the market. 9 In addition to providing information to consumers, the National Development Reform Commission, 10 which was in charge of designing the policies, and the Ministry of Finance launched in 2009 the 11 “energy-saving products and civilian-benefiting project” (Zhan et al. 2011). It covered air conditioners, 12 refrigerators, flat panel televisions, washing machines, electrical efficient lighting, energy saving and 13 new energy vehicles with the energy grades at 1 or 2 and it consisted of financial subsidies for 14 enterprises producing these products. The standard design of these financial subsidies involved the 15 government paying for the price difference of energy efficiency products and general products. The 16 manufacturers which produce the energy efficient products can get the financial subsidies directly from 17 the government (Wang et al. 2017b). 18 Before 2008, the market share of grade 1 and grade 2 air conditioners was about 5%, and about 70% of 19 all air conditioners were grade 5 (the most inefficient). Driven by the financial subsidies, the selling 20 price of the highly efficient air conditioners became competitive with that of the general air 21 conditioners. Hence, the sales of energy efficient air conditioners increased substantially, making the 22 market share of air conditioners at grade 1 and 2 to be about 80% in 2010 (Wang et al. 2017b). 23 According to the information from energy efficiency labelling management centre of China National 24 Institute of Standardisation, under the energy label system implemented 5 years ago, more than 1.5 25 hundred billion kWh power was saved by March 2010, equivalent to more than 60 million tons of 26 standard coal, 1.4 billion tons of carbon dioxide emissions, and 60 tons of sulphur dioxide emissions 27 (Zhan et al. 2011), which significantly contributed to energy saving goals of the 11th Five-Year Plan. 28 END BOX 16.7 HERE 29 Voluntary approaches and innovation 30 Voluntary approaches have a largely positive impact on innovation for those that choose to participate 31 (robust evidence, medium agreement). Research on voluntary approaches focuses on firms adopting 32 voluntary environmental management systems that can be certified based on standards of the widely 33 adopted International Standards Organisation (ISO 14001) or the European Union Environmental 34 Management and Auditing Scheme (EMAS), which is partly mandatory today. Out of 16 analyses, 70% 35 report positive innovation outcomes in terms of patents, or product and process innovation. 17% report 36 negligible impacts and 13% report negative impact. Positive innovation outcomes have been linked to 37 firms’ internal resource management practices and were found to be strengthened in firms with mature 38 environmental management systems and in the presence of other environmental regulations (He and 39 Shen 2019; Inoue et al. 2013; Li et al. 2019a). Overall, studies are concentrated in a few countries that 40 do not fully capture where environmental management systems have been actually adopted (Boiral et 41 al. 2018). There is a need for research in analyses of such instruments in emerging economies including 42 China and India, and methodologically in qualitative and longitudinal analyses (Boiral et al. 2018). 43 Competitiveness and distributional outcomes of soft instruments 44 The outcomes for performance or endorsement labels have been associated with positive 45 competitiveness outcomes (medium evidence, medium agreement). Out of 19 studies, 89% report 46 positive impact and 11% negligible impact. Although there are several studies analysing 47 competitiveness related metrics, evidence on most individual metrics is sporadic, except for housing Do Not Cite, Quote or Distribute 16-59 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 premiums. A large number of studies quantitatively assessing competitiveness find that green labels in 2 buildings are associated with housing price premium in multiple countries and regions (Fuerst and 3 McAllister 2011; Kahn and Kok 2014; Zhang et al. 2017). 32% of the studies were qualitative, 4 associating appliance labelling programs with employment and industry development (European 5 Commission 2018). There is a research gap in analyses of developing countries, and also in 6 quantitatively assessing outcomes beyond housing price premiums. 7 A few studies on the distributional outcomes of voluntary labelling programs point to positive impacts 8 (limited evidence, high agreement). All four studies focusing benefits for consumers and tenants, report 9 positive impacts (Devine and Kok 2015). Although there are benefits for utilities and other stakeholders, 10 more research is needed specifically attribute these benefits to voluntary labels rather than energy 11 efficiency programs in general. 12 Voluntary agreements are associated with positive competitiveness outcomes (medium evidence, 13 medium agreement), 14 out of 19 evaluations identified were associated with positive outcomes while 14 three were associated with negligible outcomes, and two with negative outcomes. Research found an 15 increase in perceived firm financial performance (de Jong et al. 2014; Moon et al. 2014). Studies also 16 show an association with higher exports as more environmentally conscious trade partners increasingly 17 value environmental certifications (Bellesi et al. 2005). More research is needed to develop evidence 18 on metrics of competitiveness besides firms’ financial performance, and especially in developing 19 countries. 20 Voluntary agreements are associated with a positive impact on distributional outcomes (limited 21 evidence, high agreement). Five studies, mainly using qualitative approaches, report a positive 22 association between a firm adopting an environmental management system and impacts on its supply 23 chains. There is a need for more studies with quantitative assessments and geographical diversity. 24 25 16.4.4.7 Summary of the size and direction of the evidence of all policy instrument types on 26 innovation outcomes 27 Positive impacts have been identified more frequently in some policies than in others. There is also a 28 lot of variation in the density of the literature. Developing countries are severely underrepresented in 29 the decarbonisation policy instrument evaluation literature aiming to understand the impact on 30 innovation. (high evidence, high agreement). 31 Figure 16.2 below indicates the extent to which some decarbonisation policy instruments have been 32 more or less investigated in terms of their impact on innovation outcomes as described in Table 16.9 33 above. For example, it indicates the extent to which there has been a greater focus of evaluations of the 34 impact of R&D investments, emissions trading schemes and voluntary approaches on innovation. It 35 also shows a limited amount of evidence on procurement, efficiency obligations with tradeable green 36 certificates (TGCs), building codes and auctions. Do Not Cite, Quote or Distribute 16-60 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 30 Number of evaluations 25 20 Negligible 15 Negative 10 Positive 5 0 1 Policy instrument type 2 Figure 16.2 Number of evaluations available for each policy instrument type covered regarding their 3 impact on innovation and direction of the assessment. The vertical axis displays the number of 4 evaluations claiming to isolate the impact of each policy instrument type on innovation outcomes as listed 5 in Table 16.9. The colour indicates whether each evaluation identified a positive impact on the innovation 6 outcome (blue), the existence of a negative impact (in orange), and no impact (in grey). It builds on 7 Peñasco et al (2021), Grubb et al (2021), Lilliestam et al (2021) and additional studies identified as part of 8 these reviews. TGC stands for tradeable green certificates. TWC stands for tradeable white certificates. 9 10 16.4.5 Trade instruments and their impact on innovation 11 There has been a long interest on the impact of Foreign Direct Investment (FDI) on domestic capacity 12 on innovation and on environmental outcomes. This section does not cover the much larger body of 13 evidence on the relationship between FDI and economic development and growth. 14 Overall, research indicates that trade can facilitate the entrance of new technologies, but the impact on 15 innovation is less clear (limited evidence, low agreement). A recent study indicates that for countries 16 with high environmental performance FDI has a negligible impact on environmental performance, 17 while on the lower end of the spectrum (countries with a lower environmental performance) may benefit 18 from FDI in terms of their environmental performance (Li et al. 2019b). One analysis on China links 19 FDI not just with improved environmental performance and energy efficiency but also innovation 20 outcomes in general (Gao and Zhang 2013). Other work links FDI with increased productivity across 21 firms (not just those engaged in climate-related technologies) through spill-overs (Newman et al. 2015). 22 In addition, Brandão and Ehrl (2019) indicate that productivity of the electric power industry is more 23 influenced by the transfer of embodied technology from other industries than by investments of the 24 power industry and that countries with high R&D stocks are the main sources of these international 25 technology spillovers and that the source countries may also benefit from the spillover. Do Not Cite, Quote or Distribute 16-61 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Other emerging work investigates the role of local content requirements on innovation outcomes and 2 suggests that it can lead to increased power costs (negative distributional impacts) and the domestic 3 innovation system benefits, measured by patents or exports are unclear if the policies are not part of a 4 holistic and longer lasting policy framework (Probst et al. 2020). 5 6 16.4.6 Intellectual property rights, legal framework and the impact on innovation 7 Virtually all countries around the world have instituted systems for the protection of creations and 8 inventions, known as Intellectual Property (IP) rights systems (WIPO 2021). While several types of 9 intellectual property exist – patents, copyright, design rights, trademarks, and more –, this section will 10 focus on patents, as the most relevant property right for technological innovations (World Intellectual 11 Property Organization 2008), and hence the most relevant for policy instruments in this context. 12 Patent systems aim to promote innovation and economic growth, by stimulating both the creation of 13 new knowledge and diffusion of that knowledge (high evidence, high agreement). National patent 14 systems, as institutions, play a central role in theories on national innovation systems (high evidence, 15 strong agreement). Patent systems are usually instituted to promote innovation and economic growth 16 (Nelson and Mazzoleni 1996; Machlup and Penrose 1950; Encaoua et al. 2006). Some countries 17 explicitly refer to this purpose in their law or legislation – for instance, the US Constitution states the 18 purpose of the US IP rights system to “promote the progress of science and useful arts”. Patent systems 19 aim to reach their goals by trying to strike a balance between the creation of new knowledge and 20 diffusion of that knowledge (Scotchmer and Green 1990; Devlin 2010; Anadon et al. 2016b). They 21 promote the creation of new knowledge (e.g. technological inventions) by providing a temporary, 22 exclusive right to the holder of the patent, thus providing incentives to develop such new knowledge 23 and helping parties to justify investments in research and development. They promote the diffusion of 24 this new knowledge via the detailed disclosure of the invention in the patent publication, and by 25 enabling a ‘market for knowledge’ via the trading of patents and the issuance of licenses (Arora et al. 26 2004). Although IP protections provide incentives to invest in innovation, they have the double effect 27 of restricting the use of new knowledge by raising prices or blocking follow-on innovation (Stiglitz 28 2008; Wallerstein et al. 1993). National patent systems, as institutions, feature prominently in models 29 and theories of National Innovation Systems (Edquist 1997; Klein Woolthuis et al. 2005). 30 The degree to which patent systems actually promote innovation is subject to debate. Patent protection 31 has been found to have a positive impact on R&D activities in patent-intensive industries, but this effect 32 was found to be conditional on access to finance (Maskus et al. 2019). Patents are believed to be 33 especially important to facilitate innovation in selected areas like pharmaceuticals, where investments 34 in developments and clinical trials are high, imitation costs are low, and there is often a 1:1 relationship 35 between a patent and a product, referred to as a “discrete” product industry (Cohen et al. 2000). At the 36 same time, there is an increasing body of theoretical and empirical literature that suggests that the 37 proliferation of patents also discourages innovation (medium evidence, low agreement). Theoretical 38 contributions note that a too stringent appropriability regime may greatly limit the diffusion of advanced 39 technological knowledge and eventually block the development of differentiated technological 40 capabilities within an industry, in what is called an ‘appropriability trap’ (Edquist 1997; Klein 41 Woolthuis et al. 2005). There has been a long-standing debate on the impact of patents and other IP 42 rights on innovation and economic development (Hall and Helmers 2019; Machlup 1958). Jaffe and 43 Lerner (2004) and Bessen & Meurer (2009) highlight how IP rights also hamper innovation in a variety 44 of ways. Other more specific contributions in the literature focus on specific factors. For example, 45 Shapiro (2001) discusses patent thickets, where overlapping sets of patent rights mean that those 46 seeking to commercialize new technology, need to obtain licenses from multiple patentees. Heller and 47 Eisenberg (1998) argue that a ‘tragedy of the anticommons’ is likely to emerge when too many parties 48 obtain the right to exclude others from using fragmented and overlapping pieces of knowledge, with Do Not Cite, Quote or Distribute 16-62 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 ultimately no one having the effective privilege of using the results of biomedical research. Reitzig et 2 al. (2007) describe the damaging effects of extreme business strategies employing patents, such as 3 patent trolling. 4 IP protection and enforcement in general may have different impacts on economic growth in different 5 types of countries (limited evidence, high agreement). There has been a significant degree of 6 harmonisation and cooperation between national IP systems over time. The most recent milestone is the 7 1994 WTO agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS Agreement) , 8 entered into by all members of the World Trade Organisation (WTO), and which sets down minimum 9 standards for the regulation by national governments of many forms of intellectual property as applied 10 to nationals of other WTO member nations (WTO 1994). Developing countries successfully managed 11 to include some flexibilities into TRIPS both in terms of timing of legislative reform and in terms of 12 the content of the reforms. In an attempt to understand the effects of the introduction of TRIPS, Falvey 13 et al. (2006) find that the effect of IP protection on growth is positively and significantly related to 14 growth for low- and high-income countries, but not for middle-income countries. They argue that low- 15 income countries benefit from increased technology flows, but middle-income countries may have 16 offsetting losses from the reduced scope for imitation. Note that Falvey et al. (2006) do not break down 17 their results in different technological areas and they do not focus on innovation, but instead growth. It 18 has been argued that the increasingly globalised IP regime through initiatives like the TRIPS agreement 19 will diminish prospects for technology transfer and competition in developing countries, particularly 20 for several important technology areas related to meeting sustainable development needs (Maskus and 21 Reichman 2017). 22 In principle, patent holders are not required to take their protected invention into use, and neither have 23 the obligation to allow (i.e., license) others to use the inventions in question (high evidence, high 24 agreement). Studies have shown that the way patent holders use their patent differs considerably across 25 industrial sectors: in pharmaceutics, patents are typically used to be the only producer of a certain good 26 (and obtain monopoly rents), while in industries like computers, semiconductors, and communications, 27 patents are often used to strengthen positions in cross-licensing negotiations and to generate licensing 28 income (Cohen et al. 2000; Foray 2004). There are also companies that predominantly obtain patents 29 for defensive reasons: they seek freedom to design and manufacture, and by owning a patent portfolio 30 themselves, they hope to prevent that they become the target of litigation by other patent holders (Hall 31 and Ziedonis 2001). Patents are often used strategically to impede the development and diffusion of 32 competing, alternative products, processes or services, by employing strategies known as ‘blanketing’ 33 and ‘fencing’ (Grandstrand 2000), although the research is not specific to the climate space. 34 There are notable but specific exceptions to the general principle that patent holders are not obliged to 35 license their patent to others. These exceptions include the compulsory license, Fair, Reasonable and 36 Non-discriminatory (FRAND) policies, and statement on licences of right (high evidence, high 37 agreement). While patent holders are, as stated above, in principle free to choose not to license their 38 innovation, there are three important exceptions to this. First, most national patent laws have provisions 39 for compulsory licensing, meaning that a government allows someone else to produce a patented 40 product or process without the consent of the patent holder, or plans to use the patent-protected 41 invention itself (WTO 2020). Compulsory licenses may be issued in cases of public interest or events 42 of abuse of the patent (Biadgleng 2009; World Intellectual Property Organization 2008). Compulsory 43 licensing is explicitly allowed in the WTO TRIPS agreement, and its use in context of medicine (for 44 instance to control diseases of public health importance, including HIV, tuberculosis and malaria) is 45 further clarified in the ‘DOHA Declaration’ from 2001 (Reichman 2009; WHO 2020). Second, 46 standard-setting organisations have policies to include patented inventions in their standards only if the 47 patent holder is willing to commit FRAND licensing conditions for those patents (Contreras 2015). 48 While a patent holder can still choose not to make such a commitment, by doing so, its patent is no Do Not Cite, Quote or Distribute 16-63 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 candidate anymore for inclusion in the standard. In the (many) fields where standards are of key 2 importance, it is very unusual for patent holders not to be willing to enter into FRAND commitments 3 (Bekkers 2017). Third, when a patent holder, at the time of filing at the patent office, opts for the 4 “licence of right” regime, in return for reduced patent fees, it enters into a contractual agreement that 5 obliges to license the patent to those that request it. While not all national patent systems feature this 6 regime, it is a feature present in the new European Community patent (EPO 2017), and may therefore 7 increase in importance. 8 For a discussion on the impact of IPR on international technology diffusion, see Box 16.9 in Section 9 16.5. 10 11 16.4.7 Sub-national innovation policies and industrial clusters 12 Research examining the impacts of sub-national policies on innovation and competitiveness is sporadic 13 – regional variations have been quantitatively assessed in US or China, or with case studies in these and 14 other countries. Research on wind energy in the United States, distributed PV balance of systems in 15 China, and renewable energy technologies in Italy have found that policies that incentivised local 16 demand were associated with inducing innovation, measured with patents (Fu et al. 2018; Gao and Rai 17 2019; Corsatea 2016). Different policies may have different impacts – for example, in the United States 18 state-level tax incentives and subsidies induced innovation within the state; but for renewable portfolio 19 standards policies in other states were associated with innovation, because of impact on demand, but 20 own-state policies were not (Fu et al. 2018). Research has also noted that the outcomes of policy and 21 regulation on innovation are spatially heterogenous, because of differences in local planning authorities 22 and capabilities (Song et al. 2019; Corsatea 2016). 23 Sub-national deployment policies have been associated with different impact on competitiveness 24 metrics (limited evidence, medium agreement). Research on green jobs show positive association 25 between sub-national policies and green jobs or green firms at the metropolitan level as well as the state 26 of provincial level, in both China and the United States (Yi 2013; Yi and Liu 2015; Lee 2017), while 27 others find no impact of renewable portfolio standards on green job growth in the state (Bowen et al. 28 2013). Other examples of competitiveness are in the impact of regional green industrial policy in 29 Brazil’s Rio Grande do Sul region in attracting auctioned contracts for wind energy (Adami et al. 2017) 30 or in the changes in net positive state revenues associated with removing tax incentives for wind 31 producers in Idaho in the US (Black et al. 2014). 32 Sub-national policies also directly support innovation and competitiveness through green incubators 33 and direct grants or R&D funding for local companies working on clean energy, intending to promote 34 local economic development (limited evidence, medium agreement). The literature on the impacts of 35 such policies on innovation and competitiveness is sparse. Some case studies and program evaluation 36 reports, primarily in the United States, have identified the impacts of sub-national policies on 37 competitiveness — for example, job creation from direct R&D funding in North Carolina (Hall and 38 Link 2015), perceptions for local industry development and support for follow-on financing for 39 companies receiving state-funded grants in Colorado (Surana et al. 2020b), and return on investments 40 for the state in research and innovation spending from the New York state’s energy agency (NYSERDA 41 2020). There is a general paucity of metrics on innovation and competitiveness for systematic 42 assessments of such programs in developed countries, and even more so in India and other developing 43 countries where such programs have been increasing (Surana et al. 2020a; Gonsalves and Rogerson 44 2019). 45 Although states and local governments increasingly support clean energy deployment as well as directly 46 support innovation given its link with economic development goals, there is a lack of systematic 47 research on the impacts of these policies at the subnational level. More research—both qualitative and Do Not Cite, Quote or Distribute 16-64 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 quantitative, and in both developed and developing countries—is needed to systematically develop 2 evidence on these impacts and to understand the reasons behind regional differences in terms of the 3 type of policy as well as the capabilities in the region. 4 5 16.4.8 System-oriented policies and instruments 6 Although previous sections summarised the research disentangling the role of individual policies in 7 advancing or hindering innovation (as well as impacts on other objectives), other research has tried to 8 characterise the impact of a policy mix on a particular outcome. Although the outcome studied was not 9 innovation, but diffusion (technology effectiveness is in the set of criteria outlined above), it seems 10 relevant to discuss overall findings. Reviewing renewable energy policies in nine OECD countries, 11 research concludes that, over time, a broad set of policies characterized by a ‘balance’ metric has been 12 put in place. This research also identifies a significant negative association between the balance of 13 policies in renewable energy and the diffusion of total renewable energy capacity but no significant 14 effect of the overall intensity (coded as the 46 weighted average of six indicators) on renewable capacity 15 (Schmidt and Sewerin 2019), indicating that a neutral conception of balance across all possible policies 16 may not be desirable and that policy mix intensity by itself does not explain technology diffusion. 17 A growing body of research aims to understand how different policies interact and how to characterise 18 policy mixes (del Río and Cerdá 2017; del Río 2010; Howlett and del Rio 2015; Rogge and Reichardt 19 2016). The empirical impact on the innovation outcomes is not yet discussed. A more detailed 20 discussion of this literature is located in Chapter 13. 21 An emerging stream of research in complex systems has suggested that relatively small changes in 22 policy near a possible tipping point in climate impacts in areas including changing strategies related to 23 investments in innovation, could trigger large positive societal feedbacks in the long term (Farmer et 24 al. 2019; Otto et al. 2020a). 25 26 16.5 International technology transfer and cooperation for transformative 27 change 28 This section covers international transfer and cooperation in relation to climate-related technologies, 29 “the flows of know-how, experience and equipment for mitigating and adapting to climate change 30 amongst different stakeholders” (IPCC 2000) as well as innovation to support transformative change 31 compared to the AR5 (IPCC 2014) and the SR1.5 (IPCC 2018a). This complements the discussion on 32 international cooperation on science and technology in Chapter 14. 33 This section first outlines the needs for and opportunities of international transfer and cooperation on 34 low-emission technologies. It then describes the main objectives and roles of these activities, and then 35 reviews recent institutional approaches within and outside the UNFCCC to support international 36 technology transfer and cooperation. Finally, it discusses emerging ideas on international transfer and 37 cooperation on technology, and possible modifications to support the achievement of climate change 38 and sustainable development goals, building up to Section 16.6. 39 40 16.5.1 International cooperation on technology development and transfer: needs and 41 opportunities 42 With the submission of their NDCs as part of the Paris Agreement, most developing countries are now 43 engaged in climate mitigation and adaptation. While technology is seen as one of the ‘means of Do Not Cite, Quote or Distribute 16-65 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 implementation’ of climate action, developing countries often have relatively limited technology 2 innovation capabilities, which requires them to access technologies developed in higher-income 3 countries with stronger innovation systems (Popp 2011; Binz et al. 2012; Urban 2018). In many cases, 4 these technologies require adaptation for the local context and needs (Sagar 2009; Anadon et al. 2016b), 5 and, once again, innovation capabilities are required to suitably adapt these technologies for local use 6 and also to create new markets and business models that are required for successful deployment 7 (Ockwell et al. 2015; Ockwell and Byrne 2016; Sagar 2009). This can lead to dependencies on foreign 8 knowledge and providers (Ockwell and Byrne 2016), negative impacts in terms of higher costs 9 (Huenteler et al. 2016a), and balance of payments constraints and vulnerability to external shocks 10 (Ebeling 2020). 11 The climate technology transition can also yield other development benefits, for instance better health, 12 increased energy access, poverty alleviation and economic competitiveness (Deng et al. 2018a), 13 including industrial development, job creation and economic growth (Altenburg and Rodrik 2017; 14 Porter and Van der Linde 1995; Lema et al. 2020; Pegels and Altenburg 2020) (See Section 16.6). The 15 growing complexity of technologies and global competition have made the development of a 16 technology into a globalized process involving the flow of knowledge and products across borders 17 (Koengkan et al. 2020; Lehoux et al. 2014). For instance, in production of electronics, Asian economies 18 have captured co-location synergies and dominate production and assembly of product components, 19 whereas American firms have adopted “design-only” strategies (Tassey 2014). In the context of 20 renewable energy technologies, “green global division of labour” has been observed, with countries 21 specialising in investments in R&D, manufacturing or deployment of renewables (Lachapelle et al. 22 2017). In the case of solar PV, for example, while many of the technical innovations emerged from the 23 US, Japan and China emphasized the manufacture of physical modules (Deutch and Steinfeld 2013) 24 (see also Box 16.4). 25 Such globalization of production and supply chains opens up economic development opportunities for 26 developing countries (Lema et al. 2020). At the same time, not all countries benefit from the 27 globalisation of innovation, as barriers remain related to finance, environmental performance, human 28 capabilities and cost (Egli et al. 2018; Weiss and Bonvillian 2013), with developing countries being 29 particularly disadvantaged at leveraging these opportunities. The gap in low-carbon technology 30 innovation between countries appears to have been reducing only amongst OECD countries (Yan et al. 31 2017; Du and Li 2019; Du et al. 2019) and the lower-income countries are not able to benefit as much 32 from low-carbon technologies. For instance, in the case of agriculture, Fuglie (2018) notes out that 33 international R&D spillovers seem to have benefited developed countries more than developing 34 countries. Gross et al (2018)also argue that the development timescales for new energy technologies 35 can extend up to 70 years, even within one country, and recommend that innovation efforts be balanced 36 between on the one hand commercialising already low-emission technologies in the demonstration 37 phase, and diffusing them globally, and on the other hand early-stage R&D spending. 38 Thus international cooperation on technology development and transfer can enable developing 39 countries to achieve their climate goals more effectively, while also addressing other sustainable 40 development goals, taking advantage, where possible, of the globalization of innovation and production 41 (Lema et al. 2020). Earlier assessments in the AR5 and SR1.5 have made it clear that international 42 technology transfer and cooperation could play a role in climate policy at both the international and the 43 domestic policy level (IPCC 2018b; Stavins et al. 2014; Somanathan et al. 2014) and for low-carbon 44 development at the regional level (Agrawala et al. 2014). The Paris Agreement also reflects this view 45 by noting that countries shall strengthen cooperative action on technology development and transfer 46 regarding two main aspects: 1) promoting collaborative approaches to research and development and 47 2) facilitating access to technology to developing country Parties (UNFCCC 2015). Furthermore, both Do Not Cite, Quote or Distribute 16-66 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 in literature and in UNFCCC deliberations, South-South technology transfer is highlighted (Khosla et 2 al. 2017) as a complement to the transfer of technology and know-how from North to the South. 3 This is consistent with literature that suggests that GHG mitigation in developing countries can be 4 enhanced by (1) technology development and transfer collaboration and a ‘needs-driven’ approach, (2) 5 development of the specific types of capacity required across the entire innovation chain and (3) 6 strengthening of the coordination and agendas across and between governance levels (including 7 domestic and international levels) (Upadhyaya et al. 2020; Zhou 2019; Khosla et al. 2017). 8 9 16.5.2 Objectives and roles of international technology transfer and cooperation efforts 10 International efforts involving technology transfer can have different objectives and roles. These 11 include access to knowledge and financial resources as well as promotion of new industries in both the 12 developed and recipient country (Huh and Kim 2018). Based on an econometric analysis of 13 international technology transfer factors and characteristics of Clean Development Mechanism (CDM) 14 projects, Gandenberger et al (2016) find that complexity and novelty of technologies explain whether 15 CDM project includes hardware technology transfer, and that factors like project size and absorptive 16 capacity of the host country do not seem to be drivers. Halleck Vega and Mandel (2018) argue that 17 ‘long-term economic relations’, for instance being part of a customs union, affects technological 18 diffusion between countries for the case of wind energy, and indicate that this has resulted in low- 19 income countries being largely overlooked. 20 There is some literature studying whether technology cooperation could complement or replace 21 international cooperation based on emission reductions, such as in the Kyoto Protocol, and whether that 22 would have positive impacts on climate change mitigation and compliance. A handful of papers 23 conducted game-theoretic analysis on technology cooperation, sometimes as an alternative for 24 cooperation on emission reductions, and found partially positive effects (Rubio 2017; Narita and 25 Wagner 2017; Bosetti et al. 2017; Verdolini and Bosetti 2017). However, Sarr and Swanson 26 (2017)model that, due to the rebound effect, technology development and transfer of resource-saving 27 technologies may not lead to envisioned emission reductions. 28 While technology cooperation can be aimed at emission reduction through mitigation projects, as 29 indicated above, not all cooperative actions directly result in mitigation outcomes. Overall, technology 30 transfer broadly has focused on i) enhanced climate technology absorption and deployment in 31 developing countries and ii) enhanced RD&D through cooperation and knowledge spill-overs. 32 16.5.2.1 Enhancing low-emission technology uptake in developing countries 33 Real-world outcomes in terms of low-emission technology deployment in developing countries may 34 vary significantly depending on both the nature of the international engagement and the domestic 35 context. While there have been some success in the enhancement of technology deployment through 36 technology transfer in some developing countries (de la Tour et al. 2011; Zhang and Gallagher 2016), 37 many others, and particularly least-developed countries, are lagging behind (Glachant and 38 Dechezleprêtre 2017). They indicate that this is due to the lack of participation in economic 39 globalisation and that climate negotiations could facilitate technology transfer to those countries 40 through the creation of global demand for low-emission technologies through stronger mitigation 41 targets that will result in lowering of costs and therefore enhanced technology diffusion. A broader 42 perspective presents a host of other factors that govern technology diffusion and commercialization in 43 developing countries, including, investment; social, cultural and behavioural, marketing and market 44 building; macroeconomics; and support policy (Bakhtiar et al. 2020). Ramos Mejía et al (2018) indicate 45 that the governance of low-emission technology transfer and deployment in developing countries is 46 frequently negatively affected by a mixture of well- and ill-functioning institutions, in a context of, for 47 instance, market imperfection, clientelist and social exclusive communities and patrimonial and/or Do Not Cite, Quote or Distribute 16-67 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 marketized states. Furthermore, existing interests, such as fossil fuel production, may also impede the 2 deployment of low-emission technologies, as highlighted in case studies of Vietnam and Indonesia 3 (Dorband et al. 2020; Ordonez et al. 2021). It is for such reasons that both domestic efforts and 4 international engagement are seen as necessary to facilitate technology transfer as well as deployment 5 in developing countries (Boyd 2012). The same has been seen as true in the case of agriculture where 6 the very successful international research efforts of the CGIAR (with remarkably favourable benefit- 7 cost ratios (Alston et al. 2021) were complemented by the national agricultural research systems for 8 effective uptake of high-yielding varieties of crops (Evenson and Gollin 2003). 9 One key area for underpinning effective technology uptake in developing countries relates to 10 capabilities for managing technological change that includes the capabilities to innovate, implement, 11 and undertake integrated planning. There is much research to indicate that the ability of a country’s 12 firms to adopt new technologies is determined by its absorptive capacity, which includes its own R&D 13 activities, human capacity (e.g., technical personnel), government involvement (including institutional 14 capacity), and the infrastructure in the country (Kumar et al. 1999), and that knowledge and capacity 15 are part of the ‘intangible assets’ or the ‘software’ of a firm or a country (Corsi et al. 2020; da Silva et 16 al. 2019; Ockwell et al. 2015). For sustainable development, capacity to plan in an integrated way and 17 implement the SDGs (Elder et al. 2016; Khalili et al. 2015), including using participatory approaches 18 (Disterheft et al. 2015), are conditional means of implementation. It also is argued that, if human capital 19 were at the focus of international climate negotiations as well as national climate policy, it could change 20 the political economy in favour of climate mitigation, which is needed for developing such capabilities 21 in advance to keep up with the required speed of transformation (Hsu 2017; Upadhyaya et al. 2020; 22 Ockwell et al. 2015; IPCC 2018b). Halleck-Vega et al (2018), in a global analysis of wind energy using 23 econometric analysis, lend quantitative credibility to the claim that a technology skill base is a key 24 determinant of technological diffusion. Activities to enhance capabilities include informational 25 contacts, research activities, consulting, education & training and activities related to technical facilities 26 (Huh and Kim 2018; Khan et al. 2020). 27 There are multiple studies drawing on empirical work also support this conclusion. For South-South 28 technology transfer between India and Kenya, not just technical characteristics, but also mutual learning 29 on how to address common problems of electricity access and poverty, was suggested as an important 30 condition for success (Ulsrud et al. 2018). Specifically for Africa, Olawuyi (2018)discusses the 31 capability gap in Africa, despite decades of technology transfer efforts under various mechanisms and 32 programmes of the UNFCCC. The study suggests that barriers need to be resolved by African countries 33 themselves, in particular inadequate access to information about imported climate technologies, lack of 34 domestic capacities to deploy and maintain imported technologies, the weak regulatory environment to 35 stimulate clean technology entrepreneurship, the absence or inadequacy of climate change laws, and 36 weak legal protection for imported technologies. Moreover, Ziervogel et al (2021) indicate that for 37 transformative adaptation, transdisciplinary approaches and capacity building shifting “the co-creation 38 of contextual understandings” instead of top-down transferal of existing knowledge would deliver better 39 results. Despite the understanding of the importance of the capacity issue, significant gaps still remain 40 on this front (Technology Executive Commitee 2019); see also 16.5.4). 41 16.5.2.2 Enhancing RD&D and knowledge spill-overs 42 As mentioned earlier, RD&D can aid both the development of new technologies as well as their 43 adoption for new use contexts. Therefore it is not surprising that international cooperation on RD&D is 44 identified as a mechanism to promote low-carbon innovation (see, for example, Suzuki (2015), 45 Technology Executive Committee (2021), Mission Innovation (2019)). This has resulted in a variety of 46 international initiatives to cooperate on technology in order to create knowledge spill-overs and develop 47 capacity. For example, the UNFCCC Technology Mechanism, amongst other things, aims to facilitate 48 finance for RD&D of climate technologies by helping with readiness activities for developing country Do Not Cite, Quote or Distribute 16-68 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 actors. In particular preparing early-stage technologies for a smoother transition to deployment and 2 commercialisation has been emphasised in the context of the Technology Executive Committee 3 (Technology Executive Committee 2017). There are numerous programmes, multilateral, bilateral and 4 private, that have facilitated RD&D, biased mostly towards mitigation (as opposed to adaptation) 5 activities, and many programmes that seemed to be about RD&D were in reality dialogues about 6 research coordination (Ockwell et al. 2015). There also are a variety of possible bilateral and multilateral 7 models and approaches for engaging in joint R&D (Mission Innovation 2019). An update by the 8 Technology Executive Committee (2021) reviewing good practices in international cooperation of 9 technology confirmed the conclusions of Ockwell et al(2015), and moreover highlighted that most 10 initiatives are led by the public sector, and that the private sector tended to get involved only in 11 incubation, commercialisation and diffusion phases. It also concluded that, although participation of 12 larger, higher-income developing countries seems to have increased, participation of least-developed 13 countries is still very low. 14 15 16.5.3 International technology transfer and cooperation: recent institutional 16 approaches 17 In the sections below, the literature on various categories of international technology cooperation and 18 transfer is discussed. 19 16.5.3.1 UNFCCC technology and capacity building institutions 20 Technology development and transfer are a part of the UNFCCC since its agreement in 1992 and has 21 undergone discussions and developments in the context of the international climate negotiations ever 22 since, as assessed in AR5 (Stavins et al. 2014). The support on "Technology Needs Assessment" to 23 developing countries was the first major action undertaken by the UNFCCC, and has undergone 24 different cycles of learning (Nygaard and Hansen 2015; Hofman and van der Gaast 2019). Since 2009, 25 the UNFCCC discussions on technology development and transfer have focussed on the Technology 26 Mechanism under the Cancun Agreements of 2010, which can be seen as the global climate governance 27 answer to redistributive claims by developing countries (McGee and Wenta 2014). The Technology 28 Mechanism consists of a Technology Executive Committee (TEC) and a Climate Technology Centre 29 and Network (CTCN). An independent review of CTCN evaluated it on five dimensions – relevance, 30 effectiveness, efficiency, impacts and sustainability – indicated that the organization is achieving its 31 mandate in all these dimensions, although there are some possible areas of improvement. The review 32 also specifically noted that “the lack of predictability and security over financial resources significantly 33 affected the CTCN’s ability to deliver services at the expected level, as did the CTCN’s lack of human 34 and organizational resources and the capacity of NDEs.” (Technology Executive Committee 2017). The 35 CTCN has overcome some of the limitations imposed by resource constraints by acting as a matchmaker 36 from an open-innovation perspective (Lee and Mwebaza 2020). The lack of financial sustainability of 37 the CTCN has been a recurring issue, which may potentially be resolved by deepening the linkage 38 between the CTCN and GCF (Oh 2020). In the meanwhile, the GCF is planning to establish the Climate 39 Innovation Facility to support and accelerate early-stage innovations and climate technologies through 40 the establishment of regional innovation hubs and climate accelerators as well as a climate growth fund 41 (Green Climate Fund 2020). 42 The ‘technology’ discussion has been further strengthened by the Paris Agreement, in which Article 10 43 is fully devoted to technology development and transfer (UNFCCC 2015). However, the political 44 discussions around technology continue to be characterised by viewing technology mostly as hardware 45 (Haselip et al. 2015), and relatively limited in scope (de Coninck and Sagar 2017). The workplans of 46 the Technology Executive Committee (TEC) and the CTCN do, however, indicate a broadening of the 47 perspective on technology (CTCN 2019; Technology Executive Commitee 2019). Do Not Cite, Quote or Distribute 16-69 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Since the Kyoto Protocol’s Clean Development Mechanism (CDM) has been operational, studies have 2 assessed its hypothesised contribution to technology transfer, including transfer of knowledge. Though 3 not an explicit objective of the CDM, numerous papers have investigated whether CDM projects 4 contribute to technology transfer (Michaelowa et al. 2019). The literature varies in its assessment. Some 5 find extensive use of domestic technology and hence lower levels of international technology transfer 6 (Doranova et al. 2010), while other indicate that around 40% of projects feature hardware or other types 7 of international transfer of technology (Murphy et al. 2015; Seres et al. 2009), depending on the nature 8 of technology, the host country and region (Cui et al. 2020) and the project type (Karakosta et al. 2012). 9 Although the CDM would generally be positively evaluated on the technology transfer contribution, it 10 was also regarded critically as the market-responsiveness and following of export implies a bias to 11 larger, more advanced economies rather than those countries most in need of technology transfer 12 (Gandenberger et al. 2016), although some countries have managed to correct that by directing the 13 projects, sub-nationally, to provinces with the greatest need (Bayer et al. 2016). Also, the focus on 14 hardware transfer of technology in evaluations of technology transfer under the CDM has been criticised 15 (Michaelowa et al. 2019; Haselip et al. 2015). Indeed, although many studies do go beyond hardware 16 in their evaluations (e.g. Murphy et al (2015)), the degree to which the project leads to a change in the 17 national system of innovation or institutional capacity development is not commonly assessed, or 18 assessed as limited (de Coninck and Puig 2015). 19 There is significantly less literature on capacity building under the UNFCCC, especially as it relates to 20 managing the technology transition. D’Auvergne and Nummelin (2017)(D´Auvergne and Nummelin 21 2017), in a legal analysis, indicate the nature, scope and principles of Article 11 on capacity building of 22 the Paris Agreement as being demand- and country-driven, following a needs approach, fostering 23 national, subnational and local ownership, and being iterative, incorporating the lessons learned, as well 24 as participatory, cross-cutting and gender-response. They also highlight that it is novel that least- 25 developed countries and SIDS are called out as the most vulnerable and most in need of capacity 26 building, and that it raises a “legal expectation” that all parties “should” cooperate to enhance the 27 capacity in developing countries to implement the Paris Agreement. These aspects are reflected in the 28 terms of reference of the Paris Committee on Capacity Building (PCCB) that was established in 2015 29 at the 21st Conference of the Parties (UNFCCC 2016; D´Auvergne and Nummelin 2017), which was 30 extended by five years at the 25th Conference of the Parties in 2019 (UNFCCC 2020a,b). In its work 31 plan for 2020-2024, amongst other things, it aims to “identifying capacity gaps and needs, both current 32 and emerging, and recommending ways to address them”. 33 An example of how innovative technologies combined with capacity development and institutional 34 innovation is combined in the context of adaptation to extreme weather in SIDS can be found in Box 35 16.8. 36 37 START BOX 16.8 HERE 38 Box 16.8 Capacity building and innovation for early warning systems in Small Island 39 Developing States 40 One of the areas of international cooperation on capacity building is adaptation, which has been 41 highlighted by both the Technology Executive Committee (Technology Executive Committee 2015; 42 Ockwell et al. 2015) and the Paris Committee on Capacity Building (UNFCCC 2020b) as an area where 43 capacity gaps remain, especially in Small Island Developing States (SIDS). 44 While adaptation was initially conceived primarily in terms of infrastructural adjustments to long-term 45 changes in average conditions (e.g., rising sea levels), a key innovation in recent years has been to 46 couple such long-term risk management to existing efforts to manage disaster risk, specifically 47 including early warning systems enabling early action in the face of climate- and weather-risk at much Do Not Cite, Quote or Distribute 16-70 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 shorter timescales (e.g., IPCC 2012), with potentially significant rates of return (e.g., Rogers and 2 Tsirkunov 2010; Hallegatte 2012; Global Commission on Adaptation 2019). 3 In recent years, deliberate international climate finance investments have focused on ensuring that 4 developing countries (and especially SIDS and LDCs) have access to improvements in 5 hydrometeorological observations, modelling, and prediction capacity, sometimes with a particular 6 focus on the people intended to benefit from the information produced (e.g., CREWS 2016). For 7 instance, on the Eastern Caribbean SIDS of Dominica, researchers took a community-based approach 8 to identify the mediating factors affecting the challenges to coastal fishing communities in the aftermath 9 of two extreme weather events (in particular hurricane Maria in 2017) (Turner et al. 2020). Adopting 10 an adaptive capacity framework (Cinner et al. 2018), they identified ‘intangible resources’ that people 11 relied on in their post-disaster response as important for starting up fishery, but also went beyond that 12 framework to conclude that the response ability on the part of governmental organisations as well as 13 other actors (e.g. fish vendors) in the supply chain is also a requirement for rebuilding and restarting 14 income-generating activity (Turner et al. 2020). Numerous other studies have highlighted capacity 15 building as adaptation priorities (Williams et al. 2020; Kuhl et al. 2020; Vogel et al. 2020; Basel et al. 16 2020; Sarker et al. 2020). 17 One of several helpful innovations in these efforts is impact-based forecasting (Harrowsmith et al. 18 2020), which provides forecasts targeted at the impact of the hazard rather than simply the 19 meteorological variable, enabling a much easier coupling to early action in response to the information, 20 enabling a more appropriate response afterwards. Automatic responses to warnings have also been 21 adopted in the humanitarian field for anticipatory action ahead (rather than simply in response to) 22 disasters triggered by natural hazards (Coughlan de Perez et al. 2015), resulting in a rapid scale-up of 23 such anticipatory financing mechanisms to tens of countries over the past few years, and emerging 24 evidence of its effectiveness. Still, the response is lacking in coherence and comprehensiveness, 25 resulting in calls for a more systematic evidence agenda for anticipatory action (Weingärtner et al. 26 2020). 27 END BOX 16.8 HERE 28 29 From the broader assessment above, despite limitations of available information, it is clear that the 30 number of initiatives and activities on international cooperation and technology transfer and capacity 31 building seem to have been enhanced since both the Cancun Agreements and the Paris Agreement 32 (Technology Executive Committee 2021). However, given the complexity and magnitude of the 33 requirements in terms of coverage of activities, the amount of committed funding, and effectiveness, 34 much more can be done. Some assessments of UNFCCC instruments specifically for technology 35 transfer to developing countries had indicated that functions such as knowledge development, market 36 formation and legitimacy in developing countries’ low-emission technological innovation systems 37 would need much more support to fulfil the Paris goals. (de Coninck and Puig 2015; Ockwell et al. 38 2015); such areas would benefit from continued attention, given their role in the overall climate 39 technology transition. 40 16.5.3.2 International RD&D cooperation and capacity building initiatives 41 Besides the UNFCCC mechanisms, there are numerous other initiatives that promote international 42 cooperation on RD&D as well as capacity building. Some of them are based on the notion of “mission- 43 oriented innovation policy” (Mazzucato and Semieniuk 2017; Mazzucato 2018), which shapes markets 44 rather than merely correcting market failures. 45 For instance, “Mission Innovation” (MI) is a global initiative consisting of 23 member countries and 46 the European Commission working together to reinvigorate and accelerate global clean energy 47 innovation with the objective to make clean energy widely affordable with improved reliability and Do Not Cite, Quote or Distribute 16-71 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 secured supply of energy. The goal is to accelerate clean energy innovation in order to limit the rise in 2 the global temperature to well below 2°C. The members to seek to increase public investments in clean 3 energy R&D with the engagement of private sectors, and foster international collaboration amongst its 4 members. A recent assessment shows that, although expenditures are rising, the aims are not met by 5 2020 (Myslikova and Gallagher 2020). Gross et al (2018)caution against too much focus on R&D 6 efforts for energy technologies to address climate change, including Mission Innovation. They argue 7 that given the timescales of commercialisation, developing new technologies now would mean they 8 would be commercially too late for addressing climate change. Huh and Kim (2018) discuss two 9 ‘knowledge and technology transfer’ projects that were eventually not pursued through beyond the 10 feasibility study phase due to cooperation and commitment problems between national and local 11 governments and highlight the need for ownership and engagement of local residents and recipient 12 governments. 13 The intellectual property right regime (see Box 16.9) can be an enabler or a barrier to energy transition. 14 For more background on IPR and impact on innovation, see Section 16.4.6. 15 16 START BOX 16.9 HERE 17 Box 16.9 Intellectual property rights (IPR) regimes and technology transfer 18 In the global context of climate mitigation technologies, it has been noted that technologies have been 19 developed primarily in industrialised countries but are urgently required in fast-growing emerging 20 economies (Dechezleprêtre et al. 2011). International technology transfer of such technologies can 21 primarily take place via three channels: (i) trade in goods, where technology is embedded in products; 22 (ii) direct foreign investments (FDI), where enterprises transfer firm-specific technology to foreign 23 affiliates, and (iii) patent licenses, where third parties obtain the right to use technologies. IPRs are 24 relevant for all these three channels. 25 Not surprisingly, then, the role of IPRs in international technology transfer of climate mitigation 26 technologies has been much discussed but also described as particularly controversial (Abdel-Latif 27 2015). The relationships between IP rights, innovation, international technology transfer and local 28 mitigation and adaptation are complex (Maskus 2010; Li et al. 2020; Abdel-Latif 2015) and there is no 29 clear consensus on what kind of an IPR regime will be most beneficial for promoting technology 30 transfer. 31 Several studies argue that, particularly in developing nations, the global IP regime has resulted in 32 delayed access, reduced competition and higher prices (Littleton 2008; Zhuang 2017) and that climate- 33 change-related technology transfer is insufficiently stimulated under the current IPR regime. 34 Compulsory licensing (as already used in medicine) is one of the routes proposed to repair this (Abdel- 35 Latif 2015; Littleton 2008). 36 There is little systematic evidence that patents and other IPRs restrict access to environmentally sound 37 technologies, since these technologies mostly are in sectors based on mature technologies where 38 numerous substitutes among global competitors are available (Maskus 2010). This might however 39 change in the future, for instance with new technologies based on plants, via biotechnologies and 40 synthetic fuels (Maskus 2010), for which Correa et al (2020) already find some evidence. 41 There also is literature that suggests weak IPR regimes have a “strong and negative impact on the 42 international diffusion of patented knowledge” (Dechezleprêtre et al. 2013; Glachant and 43 Dechezleprêtre 2017). Also, patents may support market transactions in technology, including 44 international technology transfer, especially to “middle-income” countries and larger developing 45 countries (Maskus 2010; Hall and Helmers 2019) but LDCs may be better served by building capacity 46 to absorb and implement technology (Sanni et al. 2016; Hall and Helmers 2010; Maskus 2010; Glachant Do Not Cite, Quote or Distribute 16-72 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 and Dechezleprêtre 2017). It is also argued that it is not even clear that the patent system as it exists 2 today is the most appropriate vehicle for encouraging international access (Sanni et al. 2016; Hall and 3 Helmers 2010; Maskus 2010; Glachant and Dechezleprêtre 2017). Given the large variation in 4 perspectives on the role of IPRs in technology transfer, there is a need for more evidence and analysis 5 to better understand if, and under what conditions, IPRs may hinder or promote technology transfer (see 6 also Technology Executive Committee (2012)). 7 In terms of ways forward to meet the challenge of climate change, different suggestions are made in the 8 context of IPRs that can help to further improve international technology transfer of climate mitigation 9 technologies, including through the TRIPS agreement, by making decisions on IPR to developing 10 countries on a case-by-case basis, by developing countries experimenting more with policies on IPR 11 protection, or through brokering or patent-pooling institutions (Littleton 2009; Dussaux et al. 2018; 12 Maskus and Reichman 2017). Others also suggests that distinctions among country groups be made on 13 basis of levels of technological and economic development, with least developed countries getting 14 particular attention (Abbott 2018; Zhuang 2017). 15 END BOX 16.9 HERE 16 17 16.5.4 Emerging ideas for international technology transfer and cooperation 18 As with the broader innovation literature (as highlighted in Section 16.3), and, in fact, drawing on such 19 literature, there has been an emergence of a greater understanding of, and emphasis on, the role of 20 innovation systems (at a national, sectoral, and technological level) as a way to help developing 21 countries with the climate technology transition (Technology Executive Committee 2015; Ockwell and 22 Byrne 2016). This has given rise to several proposals, discussed here and summarised in Figure 16.3. 23 Enhancing deployment and diffusion of climate technologies in developing countries would require a 24 variety of actors with sufficient capabilities (robust evidence, medium agreement) (Ockwell et al. 2018; 25 Kumar et al. 1999; Sagar et al. 2009). This may include strengthening existing actors (Malhotra et al. 26 2021), supporting science, technology, and innovation-based start-ups to meet social goals (Surana et 27 al 2020b), and developing entities and programs that are intended to address specific gaps relating to 28 technology development and deployment (Ockwell et al. 2018; Sagar et al. 2009). 29 There also is an increasing emphasis on the relevance of participative social innovation, local grounding 30 and policy learning as a replacement of the expert-led technological change (Kowarsch et al. 2016; 31 Chaudhary et al. 2012; Disterheft et al. 2015). Others have suggested a shift to international innovation 32 cooperation rather than technology transfer, which implies donor-recipient relationship. The notion of 33 innovation cooperation also makes more explicit the focus on innovation processes and systems 34 (Pandey et al. 2021). A broad transformative agenda therefore proposes that contemporary societal 35 challenges are complex and multi-variegated in scope and will require the actions of a diverse set of 36 actors to both formulate and address the policy, implying social, institutional and behavioural changes 37 next to technological innovations are the possible solutions (Geels 2004) (see also Cross-Chapter Box 38 12 in this chapter). 39 Several authors have proposed new mechanisms for international cooperation on technology. Ockwell 40 and Byrne (2016) argue that a role for the UNFCCC Technology Mechanism could be to support climate 41 relevant innovation-system builders (CRIBs) in developing countries, institutions locally that develop 42 capabilities that “form the bedrock of transformative, climate-compatible, technological change and 43 development”. Khan et al (2020) propose a specific variant with universities in developing countries 44 serving as ‘central hubs’ for capacity building to implement the NDCs as well as other climate policy 45 and planning instruments; they also suggest that developing countries outline their capacity building 46 needs more clearly in their NDCs. Do Not Cite, Quote or Distribute 16-73 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Building on an earlier discussion of technology-oriented and sectoral agreements (Meckling and Chung 2 2009) and the potential for international cooperation in energy-intensive industry (Åhman et al. 2017), 3 where deep emission reduction measures require transformative changes (see also Chapter 11), 4 Oberthür et al (2021) propose that the global governance for energy-intensive industry through sub- 5 sector ‘clubs’ that include governmental, private and societal actors could be effective way forward 6 (Oberthür et al. 2021). 7 Examples of emerging ideas for international cooperation on climate technology, as well as their 8 relation to the objectives and existing efforts, and in relation to the level of development of the 9 innovation system around a technology (Bergek et al. 2008; Hekkert et al. 2007) or in nations (Lundvall 10 et al. 2009) are summarized in Figure 16.3. 11 12 Figure 16.3 Examples of recent mechanisms and emerging ideas (right column) in relation to level of 13 maturity of the national or technological innovation system, objectives of international climate technology 14 transfer efforts and current mechanisms and means. Sources: (Oberthür et al. 2021; Khan et al. 2020; 15 Ockwell and Byrne 2016; Sagar 2009) 16 17 Do Not Cite, Quote or Distribute 16-74 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 16.6 Technological change and sustainable development 2 This section considers technological innovation in the broader context of sustainable development, 3 recognising that technological change happens within social and economic systems, and therefore 4 technologies are conceived and applied in relation to those systems (Grübler 1998). Simplifications of 5 complex interactions between physical and social systems and incomplete knowledge of technological 6 innovation indirect effects may systematically lead to underestimation of environmental impacts and 7 overestimation of our ability to mitigate climate change (Arvesen et al. 2011; Hertwich and Peters 8 2009). 9 In previous sections, the chapter discussed how a systemic approach, appropriate public policies and 10 international cooperation on innovation can enhance technological innovation. This section provides 11 more details on how innovation and technological change, sustainable development and climate change 12 mitigation intertwine. 13 14 16.6.1 Linking sustainable development and technological change 15 Sustainable development and technological change are deeply related (UNCTAD 2019). Technology 16 has been critical for increasing productivity as the dominant driving force for economic growth but also, 17 the concentration of technology in few hands has boosted consumption of goods and services which are 18 not necessarily aligned with sustainable development goals (Walsh et al. 2020). It has been suggested 19 that, in order to address sustainable development challenges, science and technology actors would have 20 to change their relation to policymakers (Ravetz and Funtowicz 1999) as well as the public (Jasanoff 21 2003). This has been further elaborated for the SDGs. The scale and ambition of the SDGs call for a 22 change in development patterns that require a fundamental shift in both current best practices and 23 guidelines for technological and investment decisions and in the wider socio-institutional systems 24 (UNCTAD 2019; Pegels and Altenburg 2020). This is needed as not all innovation will lead to 25 sustainable development patterns (Altenburg and Pegels 2012; Lema et al. 2015). 26 Current Sustainable Development Goals (SDG) implementation gaps reflect, to some extent, inadequate 27 understanding of the complex relationships among the goals (Skene 2020; Waiswa et al. 2019), as well 28 as their synergies and trade-offs, including how they limit the range of responses available to 29 communities and governments, and potential injustices (Thornton and Comberti 2017). These 30 relationships have been approached by focusing primarily on synergies and trade-offs while lacking the 31 holistic perspective necessary to achieve all the goals (Nilsson et al. 2016; Roy et al. 2018). 32 A more holistic framework could envisage the SDGs as outcomes of stakeholder engagement and 33 learning processes directed at achieving a balance between human development and environmental 34 protection (Gibbons 1999; Jasanoff 2003), to the extent that the two can be separated. From a science, 35 technology and innovation (STI) perspective, Fu et al (2019) distinguish three categories of SDGs. The 36 first category comprises those SDGs representing essential human needs for which inputs that put 37 pressure on sustainable development would need to be minimized. These include food (SDG 2), water 38 (SDG 6) and energy (SDG 7) resources, which continue to rely on production technologies and practices 39 that are eroding ecosystem services, hampering the realization of SDG goals 15 (land) and 14 (oceans) 40 (Díaz et al. 2019). The second are those related to governance and which compete with each other for 41 scarce resources, such as infrastructure (SDG 9) and climate action (SDG 13), which require an 42 interdisciplinary perspective. The third category are those that require maximum realization, include no 43 poverty (SDG 1), quality education (SDG 4) and gender equality (SDG 5) (Fu et al. 2019). 44 Resolving tensions between the SDGs requires adoption and mainstreaming of novel technologies that 45 can meet needs while reducing resource waste and improving resource-use efficiency, and while 46 acknowledging the systemic nature of technological innovation, which involve many levels of actors, Do Not Cite, Quote or Distribute 16-75 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 stages of innovation and scales (Anadon et al. 2016b). Changes in production technology have been 2 found effective to overcome trade-offs between food and water goals (Gao and Bryan 2017). Innovative 3 technologies at the food, water and energy nexus are transforming production processes in industrialized 4 and developing countries, such as developments in agrivoltaics, which is co-development of land for 5 agriculture and solar with water conservation benefits (Barron-Gafford et al. 2019; Schindele et al. 6 2020; Lytle et al. 2020), and other renewably powered low- to zero-carbon food, water and energy 7 systems (He et al. 2019). Silvestre and Ţîrcă (Silvestre and Ţîrcă 2019) indicate that maximising both 8 social and environmental aims is not possible, but that sustainable innovations include satisfactory 9 solutions for social, environmental and economic pillars (see Figure 16.4). 10 11 12 Figure 16.4: Considerations and typology of innovations for sustainable development (Silvestre and Ţîrcă 13 2019). 14 There is evidence that technological changes can catalyse implementation of the reforms needed to the 15 manner in which goods and services are distributed among people (Fu et al. 2019). A recently developed 16 theoretical framework based on a capability approach (CA) has been used to evaluate the quality of 17 human life and the process of development (Haenssgen and Ariana 2018). Variations of the CA have 18 been applied to exploratory studies of the link between technological change, human development, and 19 economic growth (Mayer 2001; Mormina 2019). This suggests that the transformative potential of 20 technology as an enabling condition is not intrinsic, but is assigned to it by people within a given 21 technological context. A failure to recognize and account for this property of technology is a root cause Do Not Cite, Quote or Distribute 16-76 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 of many failed attempts at techno-fixing sustainable development projects (Stilgoe et al. 2013; Fazey et 2 al. 2020). 3 The basic rationale for governance of technological change is the creation and maintenance of an 4 enabling environment for climate and SDG-oriented technological change (Avelino et al. 2019). Such 5 an environment poses high demands on governance and policy to coordinate with actors and provide a 6 direction for innovation and technological change. Cross-Chapter Box 12 illustrates how the dynamics 7 of socio-technical transitions and shifting development pathways towards sustainable development 8 offer options for policymakers and other actors to accelerate the system transitions needed for both 9 climate change mitigation and sustainable development. Governance interventions to implement the 10 SDGs will need to be operationalized at sub-national, national and global levels and support integration 11 of resource concerns in policy, planning and implementation (UNEP 2015; Williams et al. 2020). 12 13 START CCB 12 HERE 14 Cross-Chapter Box 12 Transition Dynamics 15 Authors: Anthony Patt (Switzerland), Heleen de Coninck (the Netherlands), Xuemei Bai (Australia), 16 Paolo Bertoldi (Italy), Sarah Burch (Canada), Clara Caiafa (Brazil/the Netherlands), Felix Creutzig 17 (Germany), Renée van Diemen (the Netherlands/United Kingdom), Frank Geels (United Kingdom/the 18 Netherlands), Michael Grubb (United Kingdom), Maria Figueroa (Venezuela/Denmark), Şiir Kilkiş 19 (Turkey), Jonathan Köhler (Germany), Catherine Mitchell (United Kingdom), Lars J. Nilsson 20 (Sweden), Patricia Perkins (Canada), Yamina Saheb (France/Algeria), Harald Winkler (South Africa) 21 Introduction: Numerous studies suggest that transformational changes would be required in many 22 areas of society if climate change is to be limited to 2°C warming or less. Many of these involve shifts 23 to low carbon technologies, such as renewable energy, which typically involve changes in associated 24 regulatory and social systems; others more explicitly concern behavioural shifts, such as towards plant- 25 based diets or cleaner cooking fuels, or, at the broadest level, a shift in development pathways. Chapter 26 1 establishes an analytic framework focusing on transitions, which chapters 5, 13, 14, 15 and 16 further 27 develop. In this cross-chapter box, we provide a complementary overview of the dynamics of different 28 kinds of transformational changes for climate mitigation and sustainable development. We first focus 29 on insights from socio-technical transitions approaches, and then expand to broader system transitions. 30 Dynamics of socio-technical transitions: A large literature documents the processes associated with 31 transformational changes in technology and the social systems associated with their production and use 32 (Köhler et al. 2019; Geels 2019). Transformational technological change typically goes hand-in-hand 33 with shifts in knowledge, behaviour, institutions, and markets (Markard et al. 2012, p. 956; Geels and 34 Schot 2010); stickiness in these factors often keeps society “locked-in” to those technologies already in 35 widespread use, rather than shift to new ones, even those that offer benefits (David 1985; Arthur 1994). 36 Exceptions often follow consistent patterns (Unruh 2002; Geels 2002); since AR5 a growing number 37 of scholars have suggested using these insights to design more effective climate policies and actions 38 (Geels et al. 2017). Chapter 1 (see Section 1.7 and Figure 1.6) represents technology diffusion and a 39 corresponding shift in policy emphasis as a continuous process; it is also useful to identify a sequence 40 of distinct stages that typically occur, associating each stage with a distinct set of processes, challenges, 41 and effective policies (Patt and Lilliestam 2018; Victor et al. 2019). Consistent with elsewhere in this 42 report (section 5.5.2 and SM.5.5.3 in Chapter 5, and section 16.3 in Chapter 16), Cross-Chapter Box 12 43 Figure 1 elaborates four distinct stages. Recognizing that even transformative technologies will 44 eventually be replaced with newer ones, the figure portrays these as occurring in a cycle. 45 Do Not Cite, Quote or Distribute 16-77 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 2 Cross-Chapter Box 12, Figure 1 Stages of socio-technical transition processes 3 The emergence stage is marked by experimentation, innovation in the laboratory, and demonstration in 4 the field, to produce technologies and system architectures (Geels 2005a). By its very nature, 5 experimentation includes both successes and failures, and implies high risks. Because of these risks, 6 especially in the case of fundamentally new technologies, government funding for research, 7 development and demonstration (RD&D) projects is crucial to sustaining development (Mazzucato 8 2015b). 9 The second stage is early adoption, during which successful technologies jump from the laboratory to 10 limited commercial application (Pearson and Foxon 2012). Reaching this stage is often described as 11 crossing the “Valley of Death”, because the cost/performance ratio for these new market entrants is too 12 low for them to appear viable to investors (Murphy and Edwards 2003). A key process in the early 13 adoption phase is induced innovation, a result of incremental improvements in both design and 14 production processes, and of mass-production of a growing share of key components (Grubb et al. 2021; 15 Nemet 2006). There is diversity across classes of technologies, and learning tends to occur faster for 16 technologies that are modular (Wilson et al. 2020), such as photovoltaics, and slower for those that 17 require site- or context-specific engineering, such as in the shift to low carbon materials production 18 (Malhotra and Schmidt 2020). Public policies that create a secure return on investment for project 19 developers can lead to learning associated with industry expansion (See Chapter 16, Figure 16.1); 20 typically these are economically and politically viable when they promote growth within a market 21 “niche”, causing little disruption to the mainstream market (Roberts et al. 2018). Direct support 22 mechanisms, including cross-subsidies, such as feed-in tariffs, and market quotas such as renewable 23 portfolio standards, are effective (Patt and Lilliestam 2018; Geels et al. 2017b; see Chapter 9 for 24 assessment of early adoption policies in the building sector). The value of these policies is less in their 25 immediate emissions reductions, but more in generating the conditions for self-sustaining 26 transformational change to take place as technologies later move from niche to mainstream (Hanna and 27 Victor 2021). 28 The third stage, diffusion, is where niche technologies become mainstream, with accelerating diffusion 29 rates (see Chapter 1, Sections 1.7 and 16.4), and is marked by changes to the socio-technical “regime”, 30 including infrastructure networks, value chains, user practices, and institutions. This stage is often the 31 most visible and turbulent, because more widespread adoption of a new technology gives rise to Do Not Cite, Quote or Distribute 16-78 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 structural changes in institutions and actors’ behaviour (e.g. increased adoption of smart phones to new 2 payment systems and social media), and because when incumbent market actors become threatened, 3 they often contest policies promoting the new technologies (Köhler et al. 2019). In the diffusion stage, 4 policy emphasis is shifted from financial support during the early adoption stage, towards supporting 5 regime-level factors needed to sustain, or cope with, rapid and widespread diffusion (Markard 2018). 6 These factors and policies are context specific. For example, Patt et al. (Patt et al. 2019) document that 7 the policies needed to expand residential charging networks for electric vehicles depend on the local 8 structure of the housing market. 9 The fourth stage is stabilisation, in which the new technologies, systems, and behaviours are both 10 standardized and insulated from rebound effects and backsliding (Andersen and Gulbrandsen 2020). 11 Sectoral bans on further investment in high carbon technologies may become politically feasible at this 12 point (Breetz et al. 2018; Economidou et al. 2020). The decline of previously dominant products or 13 industries can lead to calls for policy-makers to help those negatively affected, enabling a just transition 14 (Newell and Simms 2020; McCauley and Heffron 2018). Political opposition to the system 15 reconfiguration that comes with integration and stabilization can also be overcome by offering 16 incumbent actors an attractive exit strategy (de Gooyert et al. 2016). 17 Because different sectors are at different stages of low-carbon transitions, and because the barriers that 18 policies need to address are stage- and often context-specific, effective policies stimulating socio- 19 technical transitions operate primarily at the sectoral level (Victor et al. 2019). This is particularly the 20 case during early adoption, where economic barriers predominate; during diffusion, policies that 21 address regime-level factors often need to deal with cross-sectoral linkages and coupling, such as those 22 between power generation, transportation, and heating (Patt 2015; Bloess 2019; Fridgen et al. 2020). 23 The entire cycle can take multiple decades. However, later stages can go faster by building on the earlier 24 stages’ having taken place elsewhere. For example, early RD&D into wind energy took place primarily 25 in Denmark, was followed by early adoption in Denmark, Germany, and Spain, before other countries, 26 including the United States, India, and China leapfrogged directly to the diffusion stage (Lacal- 27 Arántegui 2019; Chaudhary et al. 2015; Dai and Xue 2015). A similar pattern played out for solar power 28 (Nemet 2019b). International cooperation, geared towards technology transfer, capacity and institution- 29 building, and finance, can help ensure that developing countries leapfrog to low-carbon technologies 30 that have undergone commercialization elsewhere (Adenle et al. 2015; Fankhauser and Jotzo 2018)(see 31 also Chapter 5, Box 5.9, Chapter 15, Section 15.5 and Section 16.5 in this chapter).. 32 This report contains numerous examples of the positive feedbacks in the centre of Cross-Chapter Box 33 12, Figure 1, predominantly arising during the early adoption and diffusion stages, and leading to rapid 34 or unexpected acceleration of change. Public acceptance of alternatives for meat leads to firms 35 improving the products, increasing political and economic feedbacks (Chapter 5, Section 5.4, Box 5.5). 36 Declining costs in solar and wind cause new investment in the power generation sector being dominated 37 by those technologies, leading to increased political support and further cost reductions (Chapter 6). In 38 buildings (Chapter 9) and personal mobility (Chapter 10), low-carbon heating systems and electric 39 passenger vehicles are gaining public acceptance, leading to improved infrastructure and human 40 resources, more employment in those sectors, and behavioural contagion. Some have argued that 41 technologies cross societal tipping points on account of these feedbacks (Obama 2017; Sharpe and 42 Lenton 2021). 43 Dynamics between enabling conditions for system transitions: Abson et al. (2017) argue that it is 44 possible to make use of “leverage points” inherent in system dynamics in order to accelerate 45 sustainability transitions. Otto et al. (2020b) argue that interventions geared towards the social factors 46 driving change can “activate contagious processes” leading towards the transformative changes 47 required for climate mitigation. These self-reinforcing dynamics involve the interaction of enabling 48 conditions including public policy and governance, institutional and technological innovation capacity, Do Not Cite, Quote or Distribute 16-79 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 behaviour change, and finance. For example, Mercure et al. (2018) simulated financial flows into fossil- 2 fuel extraction, and showed how investors’ taking into account transition risk in combination with 3 technological innovation would lead to the enhancement of investments in low-carbon assets and further 4 enhanced innovation. As another example, behaviour, lifestyle, and policy can also initiate demand- 5 side transitions (Chapter 5) (Tziva et al. 2020), such as with food systems (Rust et al. 2020) (Chapter 6 7, section 7.4.5), and can contribute to both resilience and carbon storage (Sendzimir et al. 2011) 7 (Chapter 16, Box 16.5). 8 In the urban context, the concept of sustainability experiments has been used to examine innovative 9 policies and practices adopted by cities that have significant impact on transition towards low-carbon 10 and sustainable futures (Bai et al. 2010; Castán Broto and Bulkeley 2013). Individual innovative 11 practices can potentially be upscaled to achieve low-carbon transition in cities (Peng and Bai 2018), 12 leading to a process of broadening and scaling innovative practices in other cities (Peng et al. 13 2019). Such sustainability experiments give rise to new actor networks, which in some cases may 14 accelerate change, and in others may lead to conflict (Bulkeley et al. 2014). As in the diffusion phase 15 in Cross-Chapter Box 12, Figure 1, contextual factors play a strong role. Examining historical 16 transitions to cycling across European cities, Oldenziel et al. (2016) found contextual factors including 17 specific configurations of actors to lead to very different outcomes. Kraus and Koch (2021) found a 18 short-term social shock – the COVID crisis – to lead to differential increases in cycling behaviour, 19 contingent on other enabling conditions. 20 Linking system dynamics to development pathways and broader societal goals: Transition 21 dynamics insights can be broadened to shifting development pathways. Development paths are 22 characterised by particular sets of interlinking regime rules and behaviours, including inertia and 23 cascading effects over time, and are reinforced at multiple levels, with varied capacities and constraints 24 on local agency occurring at each level (Burch et al. 2014)(See also Cross-Chapter Box 5 in Chapter 25 4). This is also observed by Schot and Kanger (2018), who identify a needed change in a “meta- 26 regime”, crossing sectoral lines in linking value-chains or infrastructure and overall development 27 objectives. In the context of the UN climate change regime, international cooperation can bring together 28 such best practices and lessons learnt (Pandey et al. 2021; Adenle et al. 2015). This is especially relevant 29 for developing countries, which often depend on technologies and financial resources from abroad, 30 witnessing their pace and direction influenced by transnational actors (Bhamidipati et al. 2019; 31 Marquardt et al. 2016), and benefiting little in terms of participating in high value-added activities 32 (Whittaker et al. 2020). 33 System transitions differ according to context, such as across industrialized and developing countries 34 (Ramos-Mejía et al. 2018), and within countries. Lower levels of social capital and trust negatively 35 impact niche commercialization (Lepoutre and Oguntoye 2018). In contexts of poverty and inequality, 36 stakeholders’ – including users’ – capabilities for meaningful participation are limited and transition 37 outcomes can end up marginalizing or further excluding social groups (Osongo and Schot 2017; Hansen 38 et al. 2018). Many studies of transitions in developing countries make note of the importance of 39 innovation in the informal sector (Charmes 2016, Box 5.10). Facilitating informal sector access to 40 renewable energy sources, safe and sustainable buildings, and finance can advance low-carbon 41 transitions (McCauley et al. 2019; Masuku and Nzewi 2021). Contrary, disregarding its importance can 42 result in misleading or ineffective innovation and climate strategies (Maharajh and Kraemer-Mbula 43 2010; Mazhar and Ummad 2014; de Beer et al. 2016; Masuku and Nzewi 2021). 44 Policies shifting innovation in climate-compatible directions can also reinforce other development 45 benefits, for instance better health, increased energy access, poverty alleviation and economic 46 competitiveness (Deng et al. 2018b; Karlsson et al. 2020; IPCC 2018a). Development benefits, in turn, 47 can create feedback effects that sustain public support for subsequent policies and hence help to secure 48 effective long-term climate mitigation (Meckling et al. 2015b; Geels 2014b; Schmidt and Sewerin Do Not Cite, Quote or Distribute 16-80 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 2017b; Breetz et al. 2018), increasing legitimacy of environmental sustainability actions (Hansen et al. 2 2018; van Welie and Romijn 2018; Herslund et al. 2018) and addressing negative socio-economic 3 impacts (Eisenberg 2019; McCauley and Heffron 2018; Henry et al. 2020a; Deng et al. 2018b). 4 Summary and gaps in knowledge: Strategies to accelerate climate mitigation can be most effective at 5 accelerating and achieving transformative change when they are synchronized with transition processes 6 in systems. They address technological stage characteristics, take advantage of high-leverage 7 intervention points, and respond to societal dynamics (Abson et al. 2017; Köhler et al. 2019; Geels et 8 al. 2017). Gaps in knowledge remain on how to tailor policy mixes, the interaction of enabling 9 conditions, the generalizability of socio-technical transition insights to other types of systems, and how 10 to harness these insights to better shift development pathways. 11 END CCB 14 HERE 12 13 16.6.2 Sustainable development and technological innovation: Synergies, trade-offs and 14 governance 15 16.6.2.1 Synergies and trade-offs 16 Policies that shift innovation in climate compatible directions can promote other development benefits, 17 for instance better health, increased energy access, poverty alleviation and economic competitiveness 18 (Deng et al. 2018a, see also Cross-Chapter Box 12). Economic competitiveness co-benefits can emerge 19 as climate mitigation policies trigger innovation that can be leveraged for promoting industrial 20 development, job creation and economic growth, both in terms of localizing low-emission energy 21 technologies value chains as well as of increased energy efficiency and avoided carbon lock-ins (Section 22 16.4). However, without adequate capabilities, co-benefits at the local level would be minimal, and they 23 would probably materialize far from where activities take place (Vasconcellos and Caiado Couto 2021; 24 Ockwell and Byrne 2016). Innovation and technological change can also empower citizens. Grass-roots 25 innovation promotes the participation of grass-roots actors, such as social movements and networks of 26 academics, activists and practitioners, and facilitate experimenting with alternative forms of knowledge 27 creation (UNCTAD 2019; Seyfang and Smith 2007). Examples of ordinary people and entrepreneurs 28 adopting and adapting technologies to local needs to address locally defined needs have been 29 documented in the development literature (van Welie and Romijn 2018) (See also Box 16.10). Digital 30 technologies can empower citizens and communities in decentralized energy systems contributing not 31 only to a more sustainable but also to a more democratic and fairer energy system (Van Summeren et 32 al. 2021) (See also Cross-Chapter Box 11 in this chapter, and Section 5.4 in Chapter 5). 33 Therefore, even though STI is an explicit focus of SDG 9, it is in fact an enabler of most SDGs 34 (UNCTAD 2019). Striving for synergies between innovation and technological change for climate 35 change mitigation with other SDGs can help to secure effective long-term climate mitigation, as 36 development benefits can create feedback effects that sustain public and political support for subsequent 37 climate mitigation policies (Meckling et al. 2015a; Geels 2014a, also Cross-Chapter Box 12). However, 38 innovation is not always geared to sustainable development, for instance, firms tend to know how to 39 innovate when value chains are left intact (Hall and Martin 2005), which tend to not be the case in 40 systemic transitions. 41 A comprehensive study of these effects distinguishes among “…anticipated-intended, anticipated- 42 unintended, and unanticipated-unintended consequences” (Tonn and Stiefel 2019). Theoretical and 43 empirical studies have demonstrated that unintended consequences are typical of complex adaptive 44 systems, and while a few are predictable, a much larger number are not (Sadras 2020). Even when 45 unintended consequences are unanticipated, they can be prevented through actor responses, for instance 46 rebound effects following the introduction of energy efficient technologies. Other examples of Do Not Cite, Quote or Distribute 16-81 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 unintended consequences include worse-than-expected physical damage to infrastructure and resistance 2 from communities in the rapidly growing ocean renewable energy sector (Quirapas and Taeihagh 2020), 3 and gaps between expected and actual performance of building integrated photovoltaic (BIPV) 4 technology (Boyd and Schweber 2018; Gram-Hanssen and Georg 2018). In the agricultural sector, new 5 technologies and associated practices that target the fitness of crop pests have been found to favour 6 resistant variants. Unintended consequences of digitalisation are reported as well (Lynch et al. 2019) 7 (see Cross-Chapter Box 11 in this chapter). 8 Innovation and climate mitigation policies can also have negative socio-economic impacts and not all 9 countries, actors and regions around the world benefit equally from rapid technological change 10 (UNCTAD 2019; Eisenberg 2019; McCauley and Heffron 2018; Henry et al. 2020b; Deng et al. 2018a). 11 In fact, socio-technical transitions often create winners and losers (Roberts et al. 2018). Technological 12 change can reinforce existing divides between women and men, rural and urban populations, and rich 13 and poor communities, as older workers displaced by technological change will not qualify for jobs if 14 they were unable to acquire new skills, weak educational systems may not prepare young people for 15 emerging employment opportunities, and disadvantaged social groups, including women in many 16 countries, often have fewer opportunities for formal education (UNCTAD 2019; McCauley and Heffron 17 2018). That is a risk regarding technological change for climate change mitigation, as emerging 18 evidence suggests that the energy transition can create jobs and productivity opportunities in the 19 renewable energy sector, but will also lead to job losses in fossil fuel and exposed sectors (Le Treut et 20 al. 2021). At the same time these new jobs may use more intensively high-level cognitive and 21 interpersonal skills compared to regular, traditional jobs, requiring higher levels of human capital 22 dimensions such as formal education, work experience and on-the-job training (Consoli et al. 2016). 23 Despite the empowerment potentials of decentralized energy systems, not all societal groups are equally 24 positioned to benefit from energy community policies, with issues of energy justice taking place within 25 initiatives, between initiatives and related actors, as well as beyond initiatives (van Bommel and 26 Höffken 2021; Calzadilla and Mauger 2018). 27 The opportunities and challenges of technological change can also differ within country regions and 28 between countries (e.g. Garcia-Casals et al. 2019). Within countries, Vasconcellos and Caiado Couto 29 (2021) show that, in the absence of policies and capacity building activities which promote local 30 recruiting, a significant part of total benefits of wind projects, especially high-income jobs and high 31 value-added activities, is captured by already higher income regions. Between countries, developing 32 countries usually have lower innovation capabilities, which means they need to import low-emission 33 technology from abroad and are also less able to adapt these technologies to local conditions and create 34 new markets and business models. This can lead to external dependencies and limit opportunities to 35 leverage economic benefits from technology transfer (Section 16.5.1). 36 This means that in countries below the technological frontier, the contribution of technological change 37 to climate change mitigation can happen primarily through the adoption and less through the 38 development of new technologies, which can reduce potential economic and welfare benefits from rapid 39 technological change (UNCTAD 2019). The adoption of consumer ICT technologies (Baller et al. 2016) 40 or renewable energy technologies (Lema et al. 2021) cannot bring least developed economies close to 41 the technological frontier without appropriate technological capabilities in other sectors and an enabling 42 innovation system (UNCTAD 2019; Malhotra et al. 2021; Vasconcellos and Caiado Couto 2021; Sagar 43 and Majumdar 2014; Ockwell and Mallett 2012; Ockwell et al. 2018). It has been argued widely that 44 both hard and soft infrastructure, as well as appropriate policy frameworks and capability building, 45 would facilitate developing countries engagement in long-term technological innovation and 46 sustainable industrial development, and eventually in achieving the SDGs (UNCTAD 2019; Ockwell 47 and Byrne 2016; Altenburg and Rodrik 2017). Do Not Cite, Quote or Distribute 16-82 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 16.6.2.2 Challenges to governing innovation for sustainable development 2 Dominant economic systems and centralized governance structures continue to reproduce unsustainable 3 patterns of production and consumption, reinforcing many economic and governance structures from 4 local through national and global scales (Johnstone and Newell 2018). Technological change, as an 5 inherently complex process (Funtowicz 2020), poses governance challenges (Bukkens et al. 2020) 6 requiring social innovation (Repo and Matschoss 2019) (See also Section 5.6 in Chapter 5, and Chapter 7 13). 8 Prospects for effectively governing SDG-oriented technological transformations require at a minimum 9 balanced views and new tools for securing the scientific legitimacy and credibility to connect public 10 policy and technological change in our society (Sadras 2020; Jasanoff 2018). Many frameworks of 11 governance have been proposed, such as reflexive governance (Voss et al. 2006), polycentric 12 governance (Ostrom 2010), collaborative governance (Bodin 2017), adaptive governance (Munene et 13 al. 2018) and transformative governance (Rijke et al. 2013; Westley et al. 2013) (see also Chapters 13 14 and 14). 15 A particular class of barriers to the development and adoption of new technologies comprises 16 entrenched power relations dominated by vested interests that control and benefit from existing 17 technologies (Chaffin et al. 2016; Dorband et al. 2020). Such interests can generate balancing feedbacks 18 within multi-level social-technological regimes that are related to technological lock-in, including 19 allocations of investment between fossil and renewable energy technologies (Unruh 2002; Sagar et al. 20 2009; Seto et al. 2016). 21 Weaker coordination and implementation capacity in some developing countries can undermine ability 22 to avoid trade-offs with other development objectives, like reinforced inequalities or excessive 23 indebtedness and increased external dependency, and can limit the potential of leveraging economic 24 benefits from technologies transferred from abroad (Section 16.5, Cross-Chapter Box 12). Van Welie 25 and Romijn (2018) show that in a low-income setting the exclusion of some local stakeholders from the 26 decision-making process may undermine sustainability transitions efforts. Countries with high levels of 27 inequality can be more prone to elite capture, non-transparent political decision making processes, 28 relations based on clientelism and patronage, and no independent judiciary (Jasanoff 2018), although 29 in particular contexts, non-elites manage to exert influence (Moldalieva and Heathershaw 2020). The 30 dominance of incumbents however implies that sustainable technological transitions could be achieved 31 without yielding any social and democratic benefits (Hansen et al. 2018). In the cultural domain, a 32 recurrent policy challenge that has been observed in most countries is the limited public support for 33 development and deployment of low carbon technologies (Bernauer and McGrath 2016). The 34 conventional approach to mobilizing such support has been to portray technological change as a means 35 of minimizing climate change. Empirical studies show that simply reframing climate policy is highly 36 unlikely to build and sustain public support (Bernauer and McGrath 2016). 37 Finally, there is a link between social and technological innovation; any innovation is grounded in 38 complex socio-economic arrangements, to which governance arrangements would need to respond (see 39 Sections 5.5 and 5.6, Chapter 13, and Cross-Chapter Box 12 in this chapter). Social innovation can 40 contribute to maximizing synergies and minimizing trade-offs in relation to technological innovation 41 and other innovative practices, but for this to materialize, national, regional and local circumstances 42 need to be taken into account and, if needed, changed. Even in circumstances of high capabilities, the 43 extent social innovation might help to promote synergies and avoid trade-offs is not easy to evaluate 44 (Grimm et al. 2013). Do Not Cite, Quote or Distribute 16-83 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 16.6.3 Actions that maximise synergies and minimise trade-offs between innovation and 2 sustainable development 3 Technological innovation may bring significant synergy in pursuing sustainable development goals, but 4 it may also create challenges to the economy, human well-being, and the environment (Thacker et al. 5 2019; Schillo and Robinson 2017; Walsh et al. 2020). The degree of potential synergies and trade-offs 6 among SDG differs from country to country and over time (see section 16.6.1.1). These potentials will 7 depend upon available resources, geographical conditions, development stage and policy measures. 8 Even though synergies and trade-offs related to technological innovation have received the least 9 attention from researchers (Deng et al. 2018a), literature show that higher synergy was found where 10 countries’ policies take into account the linkages between sectors (Mainali et al. 2018). For technology 11 innovation to be effective in enhancing synergies and reducing trade-offs, its role and nature in 12 production and consumption patterns, as well as in value chains and in the wider economy, requires 13 clarification. Technology ownership and control together with its current orientation and focus towards 14 productivity needs to be revised if a meaningful contribution to the implementation of the SDGs in a 15 transformative way is to be achieved (Walsh et al. 2020). Responsible innovation, combining 16 anticipation, reflexivity, inclusion and responsiveness, has been suggested as a framework for 17 conducting innovation (Stilgoe et al. 2013). Also inclusive innovation (Hoffecker 2021) could make 18 sure that unheard voices and interests are included in decision-making, and methods for this have been 19 implemented in practice (Douthwaite and Hoffecker 2017). 20 There are several examples on how to maximize synergies and avoid or minimize trade-offs when 21 bringing technological innovation to the ground. When implementing off-grid solar energy in Rwanda, 22 synergies were found between 80 of the 169 SDG targets, demonstrating how mainstreaming off-grid 23 policies and prioritising investment in the off-grid sector can realise human development and well- 24 being, build physical and social infrastructures, and achieve sustainable management of environmental 25 resources (Bisaga et al. 2021). Another example is related to wind power in Northeast of Brazil where 26 the creation of direct and indirect jobs has been demonstrated in areas where capabilities are high, as 27 well as associated improvements in wholesale and retail trade and real estate activities, though this also 28 emphasises the need for capacity development along with international collaboration projects 29 (Vasconcellos and Caiado Couto 2021). Other examples are studies raising awareness on solar energy 30 and women empowerment (Winther et al. 2018) and recycling and waste (Cross and Murray 2018). 31 Other actions with the potential to maximize synergies are those related to community or “grassroot” 32 technological innovation. The importance of the link between technological innovation and community 33 action and its contribution to sustainable development is usually underestimated, and requires further 34 research and, most importantly, its inclusion in the political agenda on sustainable development 35 (Seyfang and Smith 2007). On the other hand, when technological innovation occurs far from where is 36 implemented and participation in the production, and hence training activities of local actors is minimal, 37 co-benefits and synergies among SDGs are limited and usually far below expectations (Vasconcellos 38 and Caiado Couto 2021; Bhamidipati and Hansen 2021). Actions by policymakers that safeguard 39 environmental and social aspects can boost synergies and maximize those co-benefits (Lema et al. 40 2021). Given that technological change impacts countries, regions and social groups differently, just 41 transition policies can be designed to ensure all regions and communities are able to take advantage of 42 the energy and other transitions (Henry et al. 2020b; McCauley and Heffron 2018). 43 Box 16.10 provides insights on how a systemic approach to technological innovation can contribute to 44 reconcile synergies and trade-offs to achieve sustainable development and mitigation goals. 45 46 START BOX 16.10 HERE Do Not Cite, Quote or Distribute 16-84 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Box 16.10 Agroecological approaches: the role of local and indigenous knowledge and 2 innovation 3 Major improvements in agricultural productivity have been recorded over recent decades (FAO 2018a). 4 However, progress has also come with social and environmental costs, high levels of greenhouse gas 5 emissions and rising demand for natural resources (UNEP 2017; Bringezu 2019; UNEP 2013; Díaz et 6 al. 2019; FAO 2018a). 7 Trend analysis indicate that, of global demand for land, a large share is projected to be supplied by 8 South America, in particular the Amazon (Lambin and Meyfroidt 2011; TEEB 2018) and Chaco forests 9 (Grau et al. 2015). In developing countries, land use change for satisfying international meat demand is 10 leading to deforestation. In Brazil, the amount of GHG emitted only by the beef cattle sector represents 11 65% of the emissions of the agricultural sector and 15% of the overall emissions of the country (May 12 2019). 13 Agricultural and food systems are complex and diverse; they include traditional food systems, mixed 14 food systems and modern food systems (Pengue et al. 2018). Multiple forms of visible and invisible 15 flows of natural resources exist in global food systems (Pascual et al. 2017; IPBES 2019; TEEB 2018). 16 Technological practices, management and changes in the food chain could help adapt to climate change, 17 reduce emissions and absorb carbon in the soils, thus contributing to carbon dioxide removal (IPCC, 18 2018, 2019). A range of technologies can be implemented, from highly technological options such as 19 transgenic crops resistant to drought (González et al. 2019), salt or pesticides resistance (OECD 2011b; 20 Kim and Kwak 2020) or smart and 4.0 agriculture (Klerkx et al. 2019), to more frugal, low-cost 21 technologies such as agroecological approaches adapted to local circumstances (Francis et al. 2003; 22 FAO 2018b). These agroecological approaches are the subject of this box. 23 For developing countries, agroecological approaches could tackle both climate change challenges and 24 food security (WGII-report, Chapter 5, Box 5.10). In SIDSs, they support livelihoods to develop local 25 food value chains can promote sustainable management of natural resources, preserve biodiversity and 26 help build resilience to climate change impacts and natural disasters (FAO 2019). Other advantages of 27 agroecological practices include their adaptation to different social, economic and ecological 28 environments (Altieri and Nicholls 2017), the fact that they are physical and financial capital-extensive, 29 and are well-integrated with the social and cultural capital of rural territories and local resources 30 (knowledge, natural resources, etc.), without leading to technological dependencies (Côte et al. 2019). 31 Agroecology is a dynamic concept that has gained prominence in scientific, agricultural and political 32 discourses in recent years (Wezel et al. 2020; Anderson et al. 2021) (Chapter 7 - Agroecology (including 33 Regenerative Agriculture); Chapter 5 WGII Box 5.10). There are different agroecological approaches, 34 three of which will be briefly discussed here: agroecological intensification, agroforestry and biochar 35 use in rice paddy fields. 36 Agricultural intensification provides ways to use land, water and energy resources to ensure adequate 37 food supply while also addressing concerns about climate change and biodiversity (Cassman and 38 Grassini 2020). The term ecological intensification (Tittonell 2014) focuses on biological and 39 ecological processes and functions in agroecosystems. In line with the development of the concept of 40 agroecology, agroecological intensification integrates social and cultural perspectives (Wezel et al. 41 2015). Agroecological intensification (Mockshell and Villarino 2019) for sub-Saharan Africa aims to 42 address both employment and food security challenges (Pretty et al. 2011; Altieri et al. 2015). 43 Another example of an agroecological approach is agroforestry. Agroforestry provides examples of 44 positive agroecological feedbacks, such as ‘the regreening of the Sahel’ in Niger. The practice is based 45 on the assisted natural regeneration of trees in cultivated fields, an old method which was slowly dying Do Not Cite, Quote or Distribute 16-85 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 out but which innovative public policies (the transfer of property rights over trees from the state to 2 farmers) helped restore (Sendzimir et al. 2011). 3 Rice paddy fields are a major source of methane. Climate change impacts and adaptation strategies can 4 affect rice production and net income of rice farmers. Biochar use in rice paddy fields has been 5 advocated as a potential strategy to reduce GHG emissions from soils, enhance soil carbon stocks and 6 nitrogen retention, and improve soil function and crop productivity (Mohammadi et al. 2020). 7 Contributions of indigenous people (Díaz et al. 2019), heritage agriculture (Koohafkan and Altieri 8 2010) and peasants agroecological knowledge (Holt-Giménez 2002) to technological innovation offer 9 a wide array of options for management of land, soils, biodiversity and enhanced food security without 10 depending on modern, foreign agricultural technologies (Denevan 1995). In farming agriculture and 11 food systems, innovation and technology based on nature could help to reduce climate change impacts 12 (Griscom et al. 2017). Evidence suggests that there are benefits to integrating tradition with new 13 technologies in order to design new approaches to farming, and that these are greatest when they are 14 tailored to local circumstances (Nicholls and Altieri 2018). 15 END BOX 16.10 HERE 16 17 16.6.4 Climate change, sustainability development and innovation 18 This section gives a synthesis of this chapter on innovation and technology development and transfer, 19 connecting it to sustainable development. 20 In conjunction with other enabling conditions, technological innovation can support system transitions 21 to limit warming, help shift development pathways, and bring about new and improved ways of 22 delivering goods and services that are essential to human well-being (high confidence). At the same 23 time, however, innovation can result in trade-offs that undermine both progress on mitigation and 24 progress towards other sustainable development goals. Trade-offs include negative externalities such 25 as environmental impacts and social inequalities, rebound effects leading to lower net emission 26 reductions or even increases in emissions, and increased dependency on foreign knowledge and 27 providers (high confidence). Digitalisation, for example, holds both opportunity for emission reduction 28 and emission-saving behaviour change, but at the same time causes significant environmental, social 29 and GHG impacts (high confidence). 30 A systemic view of innovation, that takes into account the roles of actors, institutions, and their 31 interactions, can contribute to both enhanced understanding of processes and outcomes of technological 32 innovation, and to interventions and arrangements that can help innovation. It can also play a role in 33 clarifying the synergies and trade-offs between technological innovation and the SDGs. Effective 34 governance and policy, implemented in an inclusive, responsible and holistic way, could both make 35 innovation policy more effective, and avoid and minimise misalignments between climate change 36 mitigation, technological innovation, and other societal goals (medium evidence, high agreement). 37 A special feature is the dynamics of transitions. Like other enabling conditions, technological 38 innovation plays both a balancing role, by inhibiting change as innovation strengthens incumbent 39 technologies and practices, and a reinforcing role, by allowing new technologies and practices to disrupt 40 the existing socio-technical regimes (high confidence). Appropriate innovation policies can help 41 organise innovation systems better, while other policies (technology push and demand pull) can provide 42 suitable resources and incentives to support and guide these innovation systems towards societally- 43 desirable outcomes, ensure the innovations are deployed at scale, and direct these dynamics not just 44 towards system transitions for climate change mitigation, but also towards addressing other SDGs. This 45 means taking into account the full life-cycle or value chain as well as analysis of synergies and trade- 46 offs. Do Not Cite, Quote or Distribute 16-86 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Against this backdrop, international cooperation on technological innovation is one of the enablers of 2 climate action in developing countries on both mitigation and adaptation (high confidence). Experiences 3 with international cooperation on technology development and deployment suggest that such activities 4 are most effective when approached as “innovation cooperation” that engenders a holistic, systemic 5 view of innovation requirements, is done in equitable partnership between donors and recipients, and 6 develops local innovation capabilities (medium evidence, high agreement). 7 Chapter 17, in particular Section 17.4, connects technological innovation with other enabling 8 conditions, such as behaviour, institutional capacity and multi-level governance, to clarify the actions 9 that could be taken, holistically and in conjunction, to strengthen and accelerate the system transitions 10 required to limit warming to be in line with the Paris Agreement and to place countries in sustainable 11 development pathways. 12 13 16.7 Knowledge gaps 14 Filling gaps in literature availability, data collection, modelling, application of frameworks and further 15 analysis in several sectors will improve knowledge on innovation and technology development and 16 transfer, including R&D to support policy making in climate change mitigation as well as adaptation. 17 These policies and related interventions need to benefit from data and methodologies for the ex-post 18 evaluation of their effectiveness. 19 This section addresses identified knowledge gaps related to what extent developing countries are 20 represented in studies on innovation and technology development and transfer; to national contexts and 21 local innovation capacity; to potential and actual contributions of businesses; to literature emphasis on 22 mitigation; to indicators to assess innovation systems; to non-technical barriers for the feasibility of 23 decarbonisation pathways; to the role of domestic IPR policy; to digitalisation in low-emissions 24 pathways; to the compliance of Paris Agreement in regard to technology and capacity building. 25 One of knowledge gap identified when assessing the literature is on the representation of developing 26 countries in studies on innovation and technology development and transfer. This includes the 27 conceptual core disciplines of the economics of innovation, innovation systems and sustainability 28 transitions. This goes both for studies on developing countries, and for authors originating from, or 29 active in, developing countries contexts. The evidence of the impact of decarbonisation policy 30 instruments applied to developing countries or SIDS is limited. Expanding the knowledge base with 31 studies with a focus on developing countries would not only allow for testing whether the theories 32 (developed by predominantly by developed-country researchers for industrialised countries) hold in 33 developing country contexts, but also yield policy insights that could help both domestic and 34 international policymakers working on climate-related technology cooperation. 35 While a growing literature has shown how technology characteristics and complexity, national context 36 and innovation capacity can influence the capacity of a country's innovation ecosystem as a result of 37 incentive and attraction policies, more research is needed to help prioritise and design policies in 38 different national contexts while filling important knowledge gaps regarding the impact of “green” 39 public procurement, lending, “green” public banking and building code policies on innovation 40 outcomes. 41 There is also a superficial understanding of the potential and actual contributions of businesses, 42 educational institutions and socially responsible programmes, particularly in developing countries, as 43 sources of innovation and early adopters of new technologies, and a notable lack of knowledge about 44 indigenous practices. Do Not Cite, Quote or Distribute 16-87 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Besides the strong bias of literature to studies originating from and based on developed countries, 2 innovation and technology literature is also skewed to mitigation, and within mitigation to energy. 3 Literature on technology innovation for adaptation is largely missing. 4 In the area of innovation studies, data are limited on the different indicators used to assess the strength 5 of the innovation system, even for energy, including global figures on R&D and demonstration 6 spending, also for developing countries, and their effectiveness. There is also a lack of a comprehensive 7 framework and detailed data to assess the strengths of low-emission innovation systems, including 8 interactions among actors, innovation policy implementation, and strength of institutions. 9 Another gap in knowledge remains between the results from energy-climate-economy models and those 10 emerging from systems transition and sustainability transition approaches, empirical case studies, and 11 the innovation system literature. If this gap is filled, the understanding of the feasibility of 12 decarbonisation pathways in light of the many non-technical barriers to technology deployment and 13 diffusion could be improved. 14 In the field of policy instruments, existing evaluations provide insufficient evidence to assess the impact 15 of decarbonisation policy instruments on innovation, as these evaluations mainly focus on 16 environmental or technological effects. The potential positive or negative role of domestic IPR policy 17 in technology transfer to least developed countries remains unclear as the literature does not show 18 agreement. Moreover, gaps remain in impact evaluations of sub-national green industrial policies, 19 which are of growing importance. The interaction between subnational and national decarbonisation 20 policies to advance innovation would also benefit from further research, particularly in developing 21 countries. 22 The understanding of the role of digitalisation in decarbonisation pathways is lacking and needs to be 23 studied from several angles. Existing studies do not sufficiently take into account knowledge on the 24 energy impact of digital technologies, in particular the increase in energy demand by digital devices, 25 and the increase in energy efficiency. They would benefit from being technology/sector/country- 26 specific. 27 The way in which digitalisation will influence the framework conditions under which decarbonisation 28 will occur, the socio-economic and behavioural barriers influencing the diffusion of technologies in the 29 long-term scenarios and the relationship with society and its effects need to be further explored. 30 Given the implications of the digital revolution for sustainability, a better characterisation of governance 31 aspects would increase understanding of the implications and possibilities of digitalisation and other 32 GPTs for policymakers. 33 Relatedly, research (both theoretical and empirical) on the impacts of imitation, or adaptation of new 34 technological solutions invented in one region and used in other regions, could fill knowledge gaps, in 35 order to accelerate the diffusion of climate-related technologies, while taking care not to reduce the 36 incentive for inventors to invest in the search for new solutions. 37 Lastly, an independent assessment about the compliance of the Paris Agreement with regard to 38 technology and capacity building as means of implementation is starting under the Enhanced 39 Transparency Framework for action and support, where a methodology of monitoring, reporting and 40 verification are being developed. There is also a lack of analysis of the full landscape of international 41 cooperation, of what is needed to meet the objectives of the UNFCCC and the Paris Agreement, and of 42 its effectiveness. 43 Do Not Cite, Quote or Distribute 16-88 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Frequently Asked Questions (FAQs) 2 FAQ 16.1 Will innovation and technological changes be enough to meet the Paris Agreement 3 objectives? 4 The Paris Agreement stressed the importance of development and transfer of technologies to improve 5 resilience to climate change and to reduce greenhouse gas emissions. However, innovation and even 6 fast technological change will not be enough to achieve Paris Agreement mitigation objectives. Other 7 changes are necessary across the production and consumption system and the society in general, 8 including behavioural changes. 9 Technological changes never happen in a vacuum, they are always accompanied by, for instance, people 10 changing habits, companies changing value chains, or banks changing risk profiles. Therefore, 11 technological changes driven by holistic approaches can contribute to accelerate and spread those 12 changes towards the achievement of climate and sustainable development goals. 13 In innovation studies, such systemic approaches are said to strengthen the functions of technological or 14 national innovation systems, so that climate-friendly technologies can flourish. Innovation policies can 15 help respond to local priorities and prevent unintended and undesirable consequences of technological 16 change, such as unequal access to new technologies across countries and between income groups, 17 environmental degradation and negative effects on employment. 18 19 FAQ 16.2 What can be done to promote innovation for climate change and the widespread 20 diffusion of low-emission and climate-resilient technology? 21 The speed and success of innovation processes could be enhanced with the involvement of a wider 22 range of actors from the industry, research and financial communities working in partnerships at 23 national, regional and international levels. Public policies play a critical role to bring together these 24 different actors and create the necessary enabling conditions, including financial support through 25 different instruments as well as institutional and human capacities. 26 The increasing complexity of technologies requires cooperation if their widespread diffusion is to be 27 achieved. Cooperation includes the necessary knowledge flow between within and between countries 28 and regions. This knowledge flow can take the form of exchanging experiences, ideas, skills, practices, 29 among others. 30 31 FAQ 16.3 What is the role of international technology cooperation in addressing climate change? 32 Technologies that are currently known but not yet widely used need to be spread around the world, and 33 adapted to local preferences and conditions. Innovation capabilities are required not only to adapt new 34 technologies for local use but also to create new markets and business models. International technology 35 cooperation can serve that purpose. 36 In fact, evidence shows that international cooperation on technology development and transfer can help 37 developing countries to achieve their climate goals more effectively, and if this is done properly can 38 also help addressing other sustainable development goals. Many initiatives exist both regionally and 39 globally to help countries in achieving technology development and transfer through partnerships and 40 research collaboration that include developed and developing countries, with a key role for 41 technological institutions and universities. Enhancing current activities would help an effective, long- 42 term global response to climate change, while promoting sustainable development. Do Not Cite, Quote or Distribute 16-89 Total pages: 141 Final Government Draft Chapter 16 IPCC AR6 WGIII 1 Globalization of production and supply of goods and services, including innovation and new 2 technologies, may open up opportunities for developing countries to advance technology diffusion; 3 however, so far not all countries have benefited from the globalisation of innovation due to different 4 barriers such as access to finance and technical capabilities. These asymmetries between countries in 5 the globalization process can also lead to dependencies on foreign knowledge and providers. 6 Not all technology cooperation directly results in mitigation outcomes. Overall, technology transfer 7 broadly has focused on enhancing climate technology absorption and deployment in developing 8 countries as well as RD&D and knowledge spill-overs. 9 Paris Agreement also reflects this view by noting that countries shall strengthen cooperative action on 10 technology development and transfer regarding two main aspects: 1) promoting collaborative 11 approaches to research and development and 2) facilitating access to technology to developing country 12 Parties. 13 Do Not Cite, Quote or Distribute 16-90 Total pages: 141