Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation

Koretsky, Zahar (Ed.); Stegmaier, Peter (Ed.); Turnheim, Bruno (Ed.); van Lente, Harro (Ed.) Book Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation Music and Change: Ecological Perspectives Provided in Cooperation with: Taylor & Francis Group Suggested Citation: Koretsky, Zahar (Ed.); Stegmaier, Peter (Ed.); Turnheim, Bruno (Ed.); van Lente, Harro (Ed.) (2023) : Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation, Music and Change: Ecological Perspectives, ISBN 9781000831412, Routledge, London, https://doi.org/10.4324/9781003213642 This Version is available at: https://hdl.handle.net/10419/281317 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ TECHNOLOGIES IN DECLINE The central questions of this book are how technologies decline, how societies deal with technologies in decline, and how governance may be explicitly oriented towards parting with ‘undesirable’ technology. Surprisingly, these questions are fairly novel. Thus far, the dominant interest in historical, economic, sociological and political studies of technology has been to understand how novelty emerges, how innovation can open up new opportunities and how such processes may be supported. This innovation bias reflects how in the last centuries modern societies have embraced technology as a vehicle of progress. It is timely, however, to broaden the social study of technology and society: next to considering the rise of technologies, their fall should be addressed, too. Dealing with technologies in decline is an important challenge or our times, as socio-technical systems are increasingly part of the problems of climate change, biodiversity loss, social inequalities and geo-political tensions. This volume presents empirical studies of technologies in decline, as well as conceptual clarifications and theoretical deepening. Technologies in Decline presents an emerging research agenda for the study of technological decline, emphasising the need for a plurality of perspectives. Given that destabilisation and discontinuation are seen as a way to accelerate sustainability transitions, this book will be of interest to academics, students and policy makers researching and working in the areas of sustainability science and policy, economic geography, innovation studies, and science and technology studies. Zahar Koretsky is post-doctoral researcher in STS and transition studies at the French Research Institute for Agriculture, Food and the Environment (INRAE) and Gustave Eiffel University in Paris. He works on sustainability transitions, focusing on a critical take on it. His current research lies within sustainability transitions in the extractive sector. Zahar has published on the history of various declined technologies across the world and on the EU’s climate change mitigation. Peter Stegmaier is assistant professor at the Department of Technology, Policy and Society, section Science, Technology and Policy Studies (STePS) of the University of Twente, the Netherlands. Peter has worked as researcher and lecturer at the Universities of Dortmund, Hagen and Düsseldorf, carrying out at the latter two the project ‘Law as a social practice—from everyday legal methods to a legal methodology for everyday work’ (2000–2004); and as lecturer and researcher in criminal and police sociology at the Ruhr University of Bochum, focusing especially on public security institutions (2004–2006). Bruno Turnheim is INRAE Research Scientist at the Interdisciplinary Laboratory on Sciences, Innovations and Societies (LISIS, Université Gustave Eiffel, France) and Honorary Research Fellow at the Manchester Institute of Innovation Research (MIoIR, University of Manchester, UK). His research focuses on socio-technical innovation and transition processes, particularly in relation to grand societal challenges. Harro van Lente is full professor of Science and Technology Studies at the Faculty of Arts and Social Science, Maastricht University. He is one of the founding fathers of the Sociology of Expectations, which studies how representations of the future shape current socio-technical developments. He has published more than 100 journal articles, book chapters and edited volumes on technology dynamics, innovation policy and knowledge production. He is recipient of the 2018 EASST Freeman Award with the co-edited book (with Marianne Boenink and Ellen Moors), Emerging Technologies for Diagnosing Alzheimer’s Disease: Innovating with Care. TECHNOLOGIES IN DECLINE Socio-Technical Approaches to Discontinuation and Destabilisation Edited by Zahar Koretsky, Peter Stegmaier, Bruno Turnheim and Harro van Lente Cover image: L.E. Baskow www.lefteyeimages.com First published 2023 by Routledge 4 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 605 Third Avenue, New York, NY 10158 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2023 selection and editorial matter, Zahar Koretsky, Peter Stegmaier, Bruno Turnheim and Harro van Lente; individual chapters, the contributors The right of Zahar Koretsky, Peter Stegmaier, Bruno Turnheim and Harro van Lente to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. The Open Access version of this book, available at www.taylorfrancis.com, has been made available under a Creative Commons Attribution-NonCommercial- NoDerivatives 4.0 (CC-BY-NC-ND) 4.0 International license. Funded by Maastricht University. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record has been requested for this book ISBN: 978-1-032-10102-6 (hbk) ISBN: 978-1-032-10098-2 (pbk) ISBN: 978-1-003-21364-2 (ebk) DOI: 10.4324/9781003213642 Typeset in Bembo by Taylor & Francis Books CONTENTS List of illustrations vii Notes on contributors x Preface xiv Acknowledgements xvi 1 Introduction: The relevance of technologies in decline 1 Zahar Koretsky, Harro van Lente, Bruno Turnheim and Peter Stegmaier PART I Conceptual explorations 13 2 Dynamics of technological decline as socio-material unravelling 15 Zahar Koretsky 3 Destabilisation, decline and phase-out in transitions research 43 Bruno Turnheim 4 Conceptual aspects of discontinuation governance: An exploration 78 Peter Stegmaier PART II Empirical explorations 117 5 Discourses around decline: Comparing the debates on coal phase-out in the UK, Germany and Finland 119 Jochen Markard, Karoliina Isoaho and Linda Widdel vi Contents 6 Mapping the territorial adaptation of technological trajectories: The phase-out of the internal combustion engine 145 Daniel Weiss and Philipp Scherer 7 The role of alternative technologies in the enactment of (dis)continuities 167 Frédéric Goulet 8 Caring for decline: The case of 16mm film artworks of Tacita Dean 185 Dirk van de Leemput and Harro van Lente PART III Governance explorations 201 9 Implementing exnovation?: Ambitions and governance complexity in the case of the Brussels Low Emission Zone 203 Ela Callorda Fossati, Bonno Pel, Solène Sureau, Tom Bauler and Wouter Achten 10 Phase-out as a policy approach to address sustainability challenges: A systematic review 225 Adrian Rinscheid, Gregory Trencher and Daniel Rosenbloom 11 The end of the world’s leaded petrol era: Reflections on the final four decades of a century-long campaign 249 Peter Newman 12 Conclusions and continuations: Horizons for studying technologies in decline 261 Zahar Koretsky, Peter Stegmaier, Bruno Turnheim and Harro van Lente Index 273 ILLUSTRATIONS Figures 1.1 Varieties of technological decline 6 2.1 Sketching a relationship between relative geographic and temporal intensity performance of configuration, with decline in the bottom right quadrant 26 2.2 Sketching decline as a relationship between relative geographic and temporal intensity of performance of configuration and types of unravelling 30 2.3 Six ideal-type pathways of unravelling overlaid on the axes of geography-temporality (shape of the arrows are for illustration only and will differ per case) 31 5.1 Focal TIS, context systems, actor groups and struggles over decline 123 5.2 Electricity generation by source (2000–2020) 124 5.3 Discourse activity over time per country 130 5.4 Storylines and responses for each country 133 5.5 Main groups of actors for each country 134 5.6 Main actor groups and whether they were in favor of or against coal phase-out 135 5.7 Most frequent storylines and actor groups for each country 136 6.1 Analytical framework with three ICE-TIS territories and context circumstances 150 6.2 Dominant technological trajectories of the USA ICE-TIS. The circular nodes represent the patents, and the arrows depict forward citations from the cited to the citing node 154 PREFACE We started writing this book in a different world—one before the Covid-19 pandemic and before Europe saw war on its soil for the first time in the 21st century. Since that time the topic of technologies in decline has expanded its possible empirical subjects: from studying how to urgently address the climate crisis (movements to stop subsidising and to ban fossil fuels, for instance) to discontinuations of some industries during the pandemic lockdowns (for instance, aviation due to shortage of demand, food production due to a shortage of workers and the breakdown of supply chains) and war (namely, destruction in Ukraine, and, for that matter, what could be the beginning of a decline in the Russian aviation, military and IT technological fields as a result of international sanctions and an exodus of specialists). By the time this book is published, the readers will be in a better position to assess how these topics have evolved in real life. We, as editors, hope the lessons offered by this edited volume will be of use. This book started with the conference panel on ‘How are technologies abandoned?’ initiated by Zahar Koretsky, Harro van Lente and Peter Stegmaier, at the European Forum for Studies of Policies for Research and Innovation (Eu-SPRI) 2020 conference. Having received ample first signs of interest from other science and technology studies and transitions researchers, we were approached by Routledge to consider turning the panel submissions into an edited volume. We decided to accept the invitation. As the book proposal was taking shape, Bruno Turnheim got increasingly involved and became a key part of the editors’ team. We may equally count the beginnings of this book from Zahar’s PhD project on decline and phase-out (2017–2021) at Maastricht University; or from Peter’s work on the governance of discontinuation of socio-technical systems in the Open Research Area project ‘DiscGo’ (2013–2016) with research partners from France (LISIS), the United Kingdom (SPRU), the Netherlands (STePS, University of Twente), and Germany (TU Dortmund), where the discontinuation governance Preface xv framework was developed. Likewise, the book may be said to have started several years prior when Bruno published his work on destabilisation. Or when Harro started engaging with the topic of Responsible Research and Innovation (RRI) and questions of sustainability and needs. The point is, for the past decade or so the topic of declining technologies, systems and practices has been with us in some way or another, in our experiences of the material world, in the expression of social claims related to impending crises, in our conceptual thinking, and in our puzzling about how the significant variety of approaches to decline may be brought together in a meaningful way. The four of us were impressed and happy to find likeminded scholars, the contributors to this volume, who were engaging with this important topic and grappling with similar puzzles, by their eagerness and by the quality of their contributions. With this book, we hope to have contributed to the identification of a community of scholars interested in the study of technologies in decline, and we look forward to seeing how this community will develop over time, including among the readers of this book. The Editors Zahar Koretsky, Peter Stegmaier, Bruno Turnheim, Harro van Lente ACKNOWLEDGEMENTS We would like to thank the contributors for their dedication and patience during the various commenting and editing stages. Big thanks go to the participants of the 2020 Eu-SPRI panel ‘How are technologies abandoned?’ for the general positive reception and constructive discussion of the ideas and, for many of you, for following up on these ideas in this volume. Likewise, we thank the participants of the WAYS-OUT workshop in Paris, 2021, for their engagement and constructive feedback. Various institutional funding sources have supported or indirectly enabled the production of the volume. This work, partly inscribed in the WAYS-OUT research project led by Bruno Turnheim, has benefited from State assistance managed by the French National Research Agency under the ‘Programme d’Investissements d’Avenir’ under the reference ANR-19-MPGA-0010. The research for this volume has also partly been enabled through results from the Open Research Area Scheme (ORA) Grant no. 464-11-057 from NWO, DFG, ANR and ESRC within the project called ‘Governance of the discontinuation of socio-technical systems’ (DiscGo). Equal acknowledgements go to the funding from the Horizon 2020 grant for the project ‘Fostering improved training tools for responsible research & innovation’ (FIT4RRI), Grant no. 741477, as well as funding by Maastricht University. We thank our publisher, Routledge, and especially the editors Jyotsna Gurung and Annabelle Harris for picking up on this important topic, their patience with us, and all of the production work that went into this book, as well as Tom Bedford for his pertinent copyediting work. 1 INTRODUCTION The relevance of technologies in decline Zahar Koretsky, Harro van Lente, Bruno Turnheim and Peter Stegmaier The central question of this book is how technologies decline. Surprisingly, this question is fairly novel. The dominant interest in historical, economic and sociological studies of technology has been to understand how novelty emerges and how innovation can open up new opportunities. This ‘innovation bias’ in the disciplines studying technology reflects how in recent centuries modern societies have embraced technology as a vehicle of progress. Indeed, the development and use of technologies have brought remarkable improvements in health, mobility and standards of living. In the last two centuries, technologies were figured as solutions to address societal problems. Yet, in a time of growing concerns related to the challenges of climate crisis, biodiversity loss, social inequalities or geo-political tensions, technologies increasingly figure as part of the problem, too (cf. Beck 1992; Douglas 1970). Technologies that once embodied progress, such as pesticides or coal-fired power production, now embody problems and stand in the way of better directions. It is timely, therefore, to broaden the horizon of technology dynamics and the technology-society relationship: next to considering the rise of technologies, we should also consider their fall, too. These are twosides oft hesamecoinwhenitcomes toh ow the relationship to technology is constantly re-negotiated in a social context. In this volume we present some outlines for the study of technological decline. 1.1 Limits to innovation After WWII most industrialised countries adopted a techno-optimistic approach. The idea was that by stimulating scientific research and technological development, society would benefit from the boons of technology. This idea became known as the ‘linear model’ as it assumed a direct line from scientific discovery to the implementation and diffusion of technologies (Godin 2017). The techno-optimistic view of the linear model was emblematically captured in the 1933 Chicago World Fair motto: ‘Science Finds, Industry Applies, Man Conforms’. DOI: 10.4324/9781003213642-1 This chapter has been made available under a CC-BY-NC-ND 4.0 license. 2 Zahar Koretsky et al. But, of course, men and women do not readily conform. From the 1970s onwards, critical discourses about the problems of technology became more prominent. A landmark event for the public recognition that technologies can also put society at risk was the publication of the 1972 Club of Rome report on the Limits of Growth. The report indicated the depletion of resources and accumulation of pollution. It also announced that the general idea that economic growth would bring progress was misguided and not necessarily consensual. While such criticisms have changed the discourses and policies on technology, it is also clear that now, 50 years later, the problems flagged have only deepened. The climate crisis and the ongoing loss of biodiversity—as corroborated by a series of IPCC and IPBES 1 reports—indicate that widely used technologies, like the internal combustion engine, coal-fired power generation or the routine preventive use of pesticides, can pose serious threats for current and future generations. The question of technology was broadened with the issues of reducing risks and increasing democratic control—questions linking to literature on critical theory of technology, where technology is seen as neither value neutral nor universal, and opposing the privileging of technical manipulation over other relations to reality (Feenberg 2017). How to stimulate technologies while avoiding unwanted sideeffects? Can unintended effects be anticipated and avoided? 1.2 Critical discourses on technology The set of questions broadened further with the economic crises of the 1980s that saw hampered industries and painful economic reshufflings in many parts of the world. Technology appeared as the stake in intensive global competitions; national industrial policies were set up to gain a favourable position in the innovation races. As with prior techno-optimism, as if part of a cyclical pattern, this was the era of strategic research and innovation, and of significant investment programmes in ICT, biotechnology and new materials. New regional specialisations emerged under technological and competitive pressures of global capitalism. Western countries faced the social consequences of deindustrialisation in heavy industry and sought to retain competitive edges by investing in R&D for advanced technological sectors. As emerging economies in the East and South became innovation powerhouses and expanding mass consumption markets, promises of growth and prosperity provided a counterpoint to problems associated with the decline of crafts and traditional industries (e.g., hand weaving in India, cf. Mamipudi 2016). Technology and innovation remained intimately and, again, techno-optimistically, tied to economic and social prosperity, but this time there was a more widespread awareness of the downsides, and particularly the social costs of regional decline in technological races. During the last few decades, various political initiatives have been proposed to address the removal of certain technologies. Compared to efforts to stimulate technology these initiatives are modest, but there have been some successes: for fl instance, the ban of chloro uorocarbons (CFCs) under the Montreal Protocol in Introduction 3 1987, the chemical substances responsible for ozone depletion which were used, for instance, in refrigerators. Various forms of technology assessment are now being used in policy settings and in firms to evaluate the desirability of technologies. In EU research, the notion of ‘responsible research and innovation’ has become commonplace, indicating an ambition to have more public control on technologies. Currently, there is a renewed emphasis on technological sovereignty (Edler et al. 2020), to gain independence from energy sources from warring states on the occasion of the war in Ukraine, or to be less dependent on supply chains that suffered from the Covid-19 pandemic. In recent decades, environmental, health and social concerns have begun to assume more central roles in studies on technology, innovation and economic development. The UN Millennium Goals and their extension into the UN Sustainable Development Goals are important milestones signalling new long-term global orientations for economic and technological development. The rising interest in managing the concerns relating to technologies might be stemming from the economic stagnation of the global Northinthe last twodecades (Streeck 2014; Albertson 2020) and dissatisfaction with the way institutionalised decision-makers have been handling economic and environmental issues (Oreskes & Conway 2010; Wille 2010). We can also observe a fatigue from current hyper consumerist societies and the unresolved environmental concerns (Gibson- Graham 2008; Escobar 2015; Hossain 2018; de Saille et al. 2020; Hickel & Kallis 2020) coming from both rich and poor regions of the world. Both types of concerns manifest in a new style of protest, one that not only frames a problem explicitly, but also articulates the need to change systems by taking them down. The actions of Fridays for Future, Black Lives Matter or Extinction Rebellion are cases in point. More societal and environmental problems such as climate change, biodiversity loss, unhealthy lifestyles, redistributive justice, privacy breaches or the spread of fake news are galvanising a return to critical discourses about technology. Against this background, calls for more desirable alternatives (e.g., eco-innovation, responsible innovation) are being complemented by calls for deliberately discontinuing existing systems deemed undesirable. Phasing out coal and fossil fuels has, for instance, become an important priority for climate action. Similarly, we are witnessing the emergence of policy objectives and programmes seeking to shift food production systems towards pesticide-free agriculture. Problems around nuclear decommissioning have been around for decades, but haven’t yet found widely accepted solutions. One of the difficulties is that attempts to discontinue technologies tend to remain largely translated into new agendas and horizons for innovative activity without fundamentally challenging its underlying logics (e.g., the ‘green growth’ oxymoron) or established systems. Significant R&D funding is being spent to invent ways to maintain and improve lifestyles without causing deterioration of the environment. Such eco-innovation optimism is, however, struggling to deliver fully on its promises: despite significant deployment of renewable energy, electric mobility or organic agriculture in some countries, these remain a far cry from the 4 Zahar Koretsky et al. ‘fundamental system transformations’ called for (IPCC 2018; EEA 2019; UNEP 2022). Meanwhile, existing technologies and underlying systems remain relatively stable (e.g., empty passenger planes flying during pandemic) or even expand (e.g., SUVs, re-opening of coal mines in Europe due to the war in Ukraine, LNG extraction plus terminals to replace natural gas shortages), while new industries that are neither ecologically sustainable nor economically necessarily viable continue to emerge (e.g., space tourism) (Markard et al. 2021). Such examples can also show how much discontinuation has to contend with contradictory or competing rationales, interests, opportunities and framings (Turnheim 2023; Stegmaier 2023; Koretsky 2023). In short, the prominent techno-optimistic discourses are under pressure, and existing socio-technical systems, ranging from energy production, to mobility, to agri-food, are increasingly under critique. As a result, many questions come to the fore: Is it possible to do away with undesirable or unsustainable technologies? If so, how? Does this necessarily involve substitution or does it involve other shifts, too? What societal, political and industrial strategies may help to reduce our dependence on harmful technologies and socio-technical systems? Should specific products or larger systems be targeted? How can investment patterns related to harmful and polluting production be discontinued? These questions require another approach to technology: exits and divestments, destabilisation and discontinuation are high on the agenda. 1.3 Studying technology beyond innovation In the scholarly fields of innovation studies and science and technology studies, the emergence of technologies has traditionally been the focus of study. Even a decade ago Elizabeth Shove noted that ‘[w]ithin the fields of innovation studies and transitions theory, processes of emergence and stabilisation are better documented and more widely discussed than those of disappearance, partial continuity and resurrection’ (Shove 2012: 363). Yet, the attempts to abandon undesirable technologies have been hampered by insufficient insights into how such processes unfold— whether they are deliberately pursued or not. The question of how technologies decline, which we are concerned with in this book, is timely and differs markedly from earlier questions about technology in society. Of course, the recognition that technologies may be disruptive is not new. A century ago, the founder of innovation studies Joseph Schumpeter coined the phrase creative destruction to characterise the role of technical change in economies. Technologies do not just bring an accumulation of improvements, he argued, but will necessarily destabilise economic sectors, too. This still holds today: think about the woes of the postal services, which suffer from the popularity of e-mail. Before Schumpeter, Karl Marx analysed the exploitative and alienating nature of capitalism and its mobilisation of technology for this, and pointed to the fundamental disruption of social structures. The destructive character of technology, the destabilisation and eventual decline of industries or organisations, the social and Introduction 5 environmental costs of capital accumulation around technology and its use have a certain regularity and genericity that makes them observable in different geographical, sectorial or temporal settings. The main starting point for this book is the observation that (desirable) exit or reduction objectives informed by critical discourses on technology are qualitatively different from fostering desirable innovation. They involve a different kind of phenomenon, requiring different skills, different interventions and different kinds of thinking: decline is not just the reverse of innovation. Moreover, deliberate decline is likely to face resistance from significant vested interests, which may be powerful incumbents as well as more vulnerable populations and communities facing to lose significantly from the end of systems they depend on. Deliberate decline entails significant challenges, such as those associated with regulating or restricting activities and livelihoods associated with ‘undesirable’ technologies, related political contestation and struggles, but also dealing with the fact that there will be winners and losers as a result of decline. Decline is likely to be as much about setting directions and objectives as it is about managing a process and handling its aftermath—including loss (Elliott 2018). The difficulty and, at the same time, opportunity is that technologies neither persist, nor disappear into oblivion automatically: they require work to do so (Callon 1987; MacKenzie and Spinardi 1995; Russell and Vinsel 2018). Moreover, formerly established systems may still be needed for very specific purposes (DDT for vector control, special purpose incandescent light bulbs, special purpose vehicles with internal combustion engines that are fossil-fuelled). At least for a transitional period, they may leave traces that outlive the discontinuation of their active use (e.g., dealing with nuclear waste long after the disconnection and dismantling of nuclear power plants) and require dedicated infrastructures. Sometimes they also threaten to come back as zombie technology because strong interests want to push them back into the market and effectively revert phase-out programmes (e.g., the revival of coal power generation) or critical discourses are shifting (e.g., nuclear energy framed as green and CO2- neutral or dirty and life-threatening). Thus, decline is neither a linear nor an irreversible process. This is new territory for policy and research alike, and calls for revisiting concepts, methods, capabilities and means for intervention. 1.4 Perspectives and concepts This book proposes to ask what kind of processes are involved, what forms of decline can be observed, what lenses and concepts can be usefully applied and what questions remain unanswered. In this volume on technologies in decline, we seek to draw on a rich empirical base, which is diverse in terms of technologies, geographic locations and political settings. We intend to explore and use various intellectual starting points and concomitant concepts. The study of technologies in decline is necessarily interdisciplinary, drawing from multiple disciplines such as sociology, history, management and economics. Yet, this diversity can only be productive when there is also some common ground, which allows us to compare and contrast empirical findings and to connect conceptual claims. 6 Zahar Koretsky et al. Overall, the meta-theoretical lens of this book is socio-technical and we should clarify what this means, what it entails and what it requires. Firstly, ‘socio-technical’ refers to the insight that the social and the technical are deeply interwoven. They do not exist in separate domains, but are mutually embedded in tight relationships (Hughes 1983; Callon 1984; Latour 1999). As a consequence, socio-technical decline includes at least a partial dis-embedding of society-technology relationships. Secondly, a ‘socio-technical’ perspective implies that technological artefacts do not exist in and of themselves but only as a part of networks, configurations or systems. They can be seen as configurations that work, as Rip and Kemp (1998) phrased it. Consequently, technology removal or technology substitution is not the appropriate unit of analysis: technologies in decline is a matter of transformations in networks, configurations or systems. Finally, the socio-technical perspective points to the interplay of social and technological dynamics. The processes of technologies in decline then involve co-evolving social processes (political, cultural, psychological) and technical processes (in design, standardisation, manufacture, etc.). The notions and terms mobilised in this volume showcase the varying foci and interests of its various contributors. Some of the notions point to emergent and longterm processes and mechanisms, such as ‘destabilisation’. Others, such as ‘discontinuation’ or ‘phase-out’, help describe policy or policy goals andthe relatedpolicyprocesses. Using past and contemporary examples, the contributions put forward different kinds of explanations, illustrate which strategies might work and which might not, and how decisions to turn away from a questionable technology could be initiated and navigated. Judging just by the topics of the contributions and the theoretical constructs employed, we may sketch a preliminary frame of reference for various forms of technology decline, see Figure 1.1. The contributors of this volume are preoccupied with FIGURE 1.1 Varieties of technological decline Introduction 7 the particular processes of decline and adjacent concepts and how they unfold over time, their relevant characteristics, phases, trade-offs and dimensions. Two dimensions appear particularly relevant to make sense of the variety of perspectives: the scope of analysis (i.e., the perimeter of the technology, system or configuration in question) and the strategic motivations vested in a decline process (i.e., the extent to which decline is actively pursued or on the contrary a more emergent process). Of course, these dimensions are to be seen as gradients rather than binary categories, but they already allow us to qualify some archetypical foci along which technologies in decline are being thought up. Decline can range from the abandonment of a particular artefact or substance (say, a specific product model) or the erosion of a socio-technical system’s relevance and centrality, to the more purposive ban of specific substances or products, or attempts to actively discontinue entire systems. 1.5 Outline of the book We structure the volume in three Parts: from conceptual explorations, to empirical ones and, finally, to governance explorations. Table 1.1 provides an overview of the chapters, and bears witness to the significant variety of key notions deployed, analytical scale and contexts, focal context and primary research focus. The conceptual Part of the book deals with three notions. The first one is decline, or technological decline, which in Zahar Koretsky’s chapter is reconceptualised away from its colloquial, umbrella-term status in the book’s title (and this introductory chapter) to a more specific and empirically supported characterisation: a measurable trend and a socio-material process of scaling down of production and/or use of a given product or process. In the chapter, Koretsky presents an overview of literature trying to understand the mechanisms of technological decline and offers, based on it, a socio-material characterisation. From a distinct conceptual starting point, Bruno Turnheim in his chapter focuses on the notion of destabilisation next to decline and phase-out. Destabilisation is understood as an emergent process of exposure of socio-technical orders to pressures significant enough to threaten their continued existence and ‘normal’ functioning, but also strategic responses of affected actors to this exposure and changing commitment to core productive engagements. Turnheim maps the theme of destabilisation in the transitions literature and proposes a research agenda. The third notion, central for Peter Stegmaier’s chapter, is discontinuation, seen as a property of a technological trajectory in which its constituting relations become misaligned to such an extent that its distinctive character is lost, and also seen as a possible result of various permutations of distributed agency, contingency, emergence or deliberate governance. In this chapter we find a discussion of how the actions of groups of actors affect both the discontinuation of a trajectory itself and of governance practices that help stabilise it. These three notions—decline, destabilisation, discontinuation—provide lenses for the exploration of the problems around passive and active withdrawal of technologies from societies. As is discussed throughout the book and returned to in the Conclusions (Koretsky et al. 2023), these notions are interrelated and provide 2 DYNAMICS OF TECHNOLOGICAL DECLINE AS SOCIO-MATERIAL UNRAVELLING Zahar Koretsky 2.1 Introduction In both mainstream policy-making and academic literature on technologies and innovations, new technologies seem to be both more desirable and more interesting phenomena than their decline. The topic of the end of life of technologies has been periodically appearing in economics and management literature (e.g., Ayres 1987; Tushman and Rosenkopf 1992; Meckling and Nahm 2019). This literature, however, I argue, suffers from a lack of granularity in terms of timescales and unit of analysis, thus remaining somewhat abstract regarding technological decline. Furthermore, typically, processes of decline tend to be conflated with substitution, e.g. in literature on disruptive innovation (e.g., Ayres 1987; Yu and Hang 2010; Taylor and Taylor 2012) and in climate models (as critiqued by Anderson and Peters (2016)). Decline is also seen as an unappealing topic altogether (Smith, Voß and Grin 2010; Stegmaier, Kuhlmann and Visser 2014) because of perceived connotations with failure and a greater interest in markets and competition (Polanyi 1944; McCarraher 2019; Vinsel and Russell 2020) as the nexus of both social life and technological development. In this chapter, 1 in contrast with prior economic and political economic literature on decline (see Stegmaier 2023), I qualify technological decline as reducing production and/or use of a given technology to a niche degree or to complete abandonment. Technological decline can be a result of emergent or coordinated processes or (more realistically) a mix of both. When a result of phase-out or discontinuation (see Stegmaier 2023), decline is an effective policy intervention which has proven instrumental for stopping wide exposure to carcinogenic DDT gas (Levain et al. 2015) and ozone-depleting aerosols (Simmonds et al. 1993), improving the energy efficiency of indoor lighting with the ban on the incandescent light bulb (Kierkegaard 2014; Stegmaier, Visser and Kuhlmann 2021) and DOI: 10.4324/9781003213642-2 This chapter has been made available under a CC-BY-NC-ND 4.0 license. 16 Zahar Koretsky more. Many countries have instigated the decline of coal and nuclear power, and policies promoting decline increasingly appear on political agendas as countries are faced with the dangerous mid- and short-term prospects of climate crisis (Rogge and Johnstone 2017). 2 Decline, however, can and will generate uncertainties in economies, environments, labour relations, national security, democracy and more (Lieu et al. 2019; Turnheim 2023). The literature, to date, has not focused on explaining why and how this occurs, nor how to navigate it. An elaborated conceptualisation of the processes involved with decline is needed, in which there would be a higher level of granularity of analysis (Murmann and Frenken 2006) than in existing literature. Thus, in this chapter, I seek an answer to the question: What are the key mechanisms of the socio-material dynamics of technological decline? When answering the research question I will conduct a theoretical comparison of three cases, and formulate conceptualisations by way of middle-range theorising. The idea of a middle-range type of theorising originates with Merton (1949). It implies an analytical focus on a specific phenomenon, rather than on the social system broadly, as in a “grand theory”. Such a narrower focus allows middle range theories to close in on the observed phenomenon, while being abstract enough to explain the observed variations between its instances across cases. According to Merton (1949), the benefit of middle range theories is that they are not derived from a larger theory of social systems (though they may be consistent with them); rather, their starting point is empirical observation and basic generalisation, which are then complexified after subsequent rounds of observation and abstraction of patterns. The importance of middle range theory has been recognised in science and technology studies (STS) and transition studies (e.g., Frickel and Moore 2006; Geels 2007; Hamlin 2007; Wyatt and Balmer 2007) as part of ongoing self-reflection on own societal value. Frickel and Moore (2006) call STS scholarship to refocus from critique to theorising, and Wyatt and Balmer (2007) ask whether STS has prematurely given up on theorising altogether in its scepticism of generalising. More recently, middle range theorising has been picked up in studies of healthcare technology implementation (Lyle 2021), urban transitions (Mora et al. 2021), social innovation (Pel et al. 2020) and sustainability transitions (Geels 2018), among others. I aim to contribute to this trend and formulate an approach helpful for future analyses of technological decline. Methodologically, I follow Christensen’s (2006) approach of descriptive and predictive theory building. Christensen suggests to start inductively, with a careful description and documentation of the observed phenomena, and to continue with formulating abstractions to rise above the messy empirical detail and to understand key operations of the phenomena. This, he asserts, makes it possible to inductively classify the abstractions into a typology to simplify and organise the studied phenomena. The final task is to identify associations between the category-defining attributes of the phenomena and the observed outcomes. Similarly, Eisenhardt and Graebner (2007) argue that the theory-building process occurs via recursive cycling Dynamics of technological decline as socio-material unravelling 17 among rich empirical data, emerging theory and extant literature. The hope is that the result will be able to produce theory that is accurate, parsimonious, interesting and testable (Eisenhardt and Graebner 2007). Following this logic, I start with a literature review in section 2.2 and formulate a conceptual framework in section 2.3 to analyse and compare empirical cases of decline in section 2.4. Based on this analysis, in section 2.5 I discuss the outlines of a middle-range theory of the dynamics of technological decline, and in the concluding section 2.6 I make final remarks and propose possible future research directions. 2.2 Decline in theories of technological change In this section I will outline some of the most prominent theories 3 in the adjacent fields of transition studies and STS on technological change that offer insightful groundwork for the study of decline, even if they tend to not discuss decline explicitly. A more detailed review can be found in Koretsky (2023). To conduct this theoretical review I searched the Web of Science for the terms “technology” in combination with (variants of) “decline”, “destabilisation”, “control”, “discontinuity”, “dismantling”, “convergence”, “extinction”, “loss”, “disappearance”, “death”, “abandonment”, “termination” and “phase out”. I compiled a list of literature of the first 600 most cited items, and manually screened their titles and abstracts for relation to the topic of the chapter, after which 43 relevant entries remained. The number of dropped entries was so high because the search terms are rather popular words found in articles that had nothing to do with the topic of technological decline. In fact, additional searches using the snowballing method, where I identified key sources and bodies of literature with the help of other literature reviews and research agendas (e.g., Köhler et al. 2017, 2019; Kanger, Sovacool and Noorkõiv 2020; Rosenbloom and Rinscheid 2020; Turnheim and Sovacool 2020), revealed more entries: namely, 107. The final list included exactly 150 academic sources. As I closely read them, I used an Excel table to note if and how the authors define and address technology, technological change, and technological decline and adjacent concepts. The literature searches were conducted in 2018 and 2022. The literature shares conceptualisations of the object of analysis and key mechanisms of change. First, configurations (Rip and Kemp 1998; Shove, Pantzar and Watson 2012), networks (Callon 1984; Latour 1999) and systems (Hughes 1983; Geels 2002; Geels and Schot 2007; Hekkert et al. 2007; Bergek et al. 2008) appear as focal entities in the analysed literature. While understood and used quite differently, these terms stem from one ontological source—a relational understanding of the world. In these three concepts, the emphasis is on heterogeneous ensembles of artefacts, infrastructures, individuals, scientific knowledge, cultural categories, cultural/symbolic meaning, consumption patterns, industry structures, natural resources, markets, and norms and laws (Geels 2004; Shove, Pantzar and Watson 2012; Hess and Sovacool 2020). Here, society and technology, humans 18 Zahar Koretsky and things are “mutually constitutive and hopelessly mangled” (Garud and Gehman 2012: 983), i.e. co-constructed or co-produced, “with no single dimension dictating change by itself” (Sovacool and Hess 2017: 31). In this literature decline tends to be understood as break of reproduction and of ties within and shrinkage of the network/system/configuration. The concepts of configuration, network and system inform us that it could be productive to look for their disruptions when tracing and conceptualising decline. A lot of attention in literature is given to meanings and sense-making, drawing from symbolic interactionism and structuration theory (Giddens 1984). The key observation is that there is a dichotomous relationship between structures of rules and meanings on the one hand and, on the other hand, the agency of actors, who are free to interpret and enact these structures and rules. Bijker et al. (1987) discuss these interpretations with their concept of “interpretive flexibility”. Interpretive flexibility foregrounds that scientific knowledge and technological design require a negotiation between interpretations and the concomitant social processes. Van Lente and Rip (1998) further argue that expectations and promises structure actions: researchers, engineers and firms propose technology options and claim promises. When accepted, such promises are typically taken as requirements to be fulfilled. Thus, in transitions theory, expectations, promises and symbolic/cultural meanings of technology are performative forces of a socio-technical system (Geels 2002; Geels and Kemp 2007). They must, then, play a key role in decline. A key mechanism of change in the literature is the co-evolution and co-design of the elements within the configuration, network and system. Material objects are often key here (Latour 1992; Bijker 1995; David and Schulte-Römer 2021). According to Geels (2020), STS scholarship has brought the discussion of the pervasive influence of material and technological dimensions to the mainstream sociological discussion about agency and structure. Callon (1987) famously suggested that technology is a way of studying “society in the making”, and Latour (1991) saw technology as “society made durable”, arguing that artefacts provide ways of anchoring new routines and practices (Geels 2020). In synthesising approaches such as the MLP or TIS materiality is included as infrastructures and resources. If materiality plays as big of a role in change as society, it must also be traced to study decline. Next, in the literature, networks, systems and configurations are typically seen as dynamically, actively and continually forming and reforming according to the intentionality of human actors. They are constrained by an accumulated “momentum”, inertia or “frame”. In ANT this (re)formation work is referred to as performance, in the sense of reproduction, “making and remaking groups” (Latour 2005: 35). Similarly, Law writes about the processes of “heterogeneous engineering” in which actors try to “maintain some degree of stability in the face of the attempts of other entities or systems to dissociate them into their component parts” (Law 1987: 248). Some authors in theories of practice also refer to performance, seeing it as the reproduction of the associations between material, cognitive and symbolic dimensions of a practice (Shove, Pantzar and Watson 2012; Welch and Dynamics of technological decline as socio-material unravelling 19 Warde 2015). For studies of decline it might be productive to mobilise the concept of performance to partly describe the dynamics of the configuration. An important process in performance is learning, or creation, retention and transfer of skills, knowledge and competences. Learning is emphasised in transition theory as an indicator of success of an innovation, and is something that is protected in niches (Loorbach, Frantzeskaki and Avelino 2017). In STS, learning is involved in processes of appropriation and domestication of technology (Sørensen 2006). Learning involves processes of reframing that may result in a change of perspective among stakeholders on the problem, upon which they will try to collectively find a solution (Grin, Rotmans and Schot 2010). In practice theory, activities related to learning and other forms of competences co-constitute configurations, together with materials and meanings (Shove, Pantzar and Watson 2012). They must be key in decline as well. Prior literature already started pushing these shared concepts of configuration/ system/network, co-evolution, sense-making, materiality, performance, competences into analyses of decline. Much of the literature discusses decline of a configuration as processes of delegitimation or competition with other configurations, to which the configuration may readjust (reconfigure) or “drift” into a declined, “dormant” state. Such mechanisms are discussed across studies of industries and firms in LTS, technology life cycle and regime destabilisation literature, TIS, social practices and discontinuation (Table 2.1). Case studies of dissociations, disconnections and de-alignments have been emerging lately with work on withdrawal. There, more attention is being paid to framing and creation of competing configurations. Other approaches, such as MLP and SCOT, have not yet developed a body of empirical work on decline and remain somewhat theoretical in their descriptions of decline mechanisms. Overall, I draw five insights from the reviewed literature for studies of decline: � Decline is a systems process constituted by technical and sense-making dynamics. � These technical and sense-making dynamics tend to involve a weakening of an element of the system (i.e., it becomes less stable, reliable, resilient, flexible compared to its previous condition). � In addition, system connections may weaken, too. � When declining, systems shrink and their role in the environment decreases. � Due to an emergence bias that has dominated studies of technological change for so long, but also due to differing units of analysis and of observation (with, I would claim, a superfluous focus on industry and firms), there is not enough empirical granularity in existing studies of technological decline (yet). Based on these insights, I claim that a good characterisation and middle-range theorisation of the dynamics of technological decline will need to meet the following criteria. They will need to: 1) be grounded in empirical studies of detailed granularity of analysis (Murmann and Frenken 2006), 2) appreciate the complexity and messiness of the socio-material dynamics of decline (i.e., sense-making, materiality, competences, 20 Zahar Koretsky expectations and cultural meanings), and 3) appreciate the developments outside of market and industry, as not the only arenas of socio-technical change. Based on the insights from the existing literatures on technology and technological change, in the next section I formulate a conceptual framework to structure the analysis of the phenomenon of technological decline. 2.3 A conceptual framework and methodology to study the dynamics of decline From the literature on technological change it appears that the basic starting point of an analysis of technologies is the acknowledgment of a strong interrelatedness of artefacts and society, economy and culture, institutions and practices. In other words, the association of material and socio-cognitive entities and their relations in a configuration are often the unit of analysis, and so they are in the present chapter. Within this unit of analysis units of observation need to be established, 4 without which the configuration will be very difficult to study with a high degree of granularity. In an attempt to avoid a critique that analysing the intertwined socio-technical dynamics amounts to concluding that “everything is in everything” (Sovacool and Hess 2017: 15 quoting Gingras 1995), the units of observation need to be analytically distinguished from each other and analytically separated to the extent that this is possible. I draw from STS literature that analytically separates the “strands” of the entangled socio-technology: Shove and colleagues (2012) who discuss materials, meanings and forms of competence; literature on domestication (e.g., Lie and Sørensen 1996; Sørensen 2006) where symbolic, practical and cognitive work is distinguished; MacKenzie and Spinardi (1995) where “tacit knowledge, control over materials, and the translation of interests form… a necessary three-sided approach to… uninvention” (MacKenzie and Spinardi 1995: 88); and Latour and Callon, who distinguish between human and non-human actors (Callon 1984; Latour 1999). Borrowing the terminology used by Shove and colleagues (2012), I thus operationalise the “strands” that coconstitute the configuration as materials, meanings and forms of competence.Whent alking about these I will mean, respectively: � objects, tools, hardware, infrastructures, production facilities, resources � laws, rules, public discourses, competing narratives of supporters and opponents before and after the decline and changes in cultural meanings and regulatory regimes they may lead to (or have led to) � tacit, codified and other knowledge on design, manufacture and use, as well as labour relations 5 (see Table 2.1). Such separation into three strands is an analytical move and not an ontological claim, and a simplification that I commit to as an operationalisation strategy (cf. Klein and Kleinman 2002; Geels 2007). The three strands are not monolithic and are themselves networks of different entities that co-exist and compete (Callon’s “punctualisation” (1984)). The dynamics Unit of analysis (the ‘strands’) Units of observation Materials Objects, tools, hardware, infrastructures, production facilities, material resources Meanings Public discourses, narratives, laws Forms of competence Tacit, codified and other knowledge on design, manufacture and use, labour relations Table reprinted from Koretsky (2022b) TABLE 2.1 Units of analysis and units of observation in the conceptual framework Dynamics of technological decline as socio-material unravelling 21 within the strands can be expected to be driven by the same rules as other networks, e.g. they are dynamically stable as long as there is alignment (Goulet 2021) and they can become unstable if there is too much internal contestation. I will consider that meanings, competences and materials are associated, entangled in a configuration as long as the configuration is performed, i.e. the technology is participating in manufacture or routine use. Every time the given technology is manufactured or used, the associations are reproduced and entangled tighter. This is a state of continuity, dynamic stability and the absence of decline. The reverse processes I propose to term “unravelling”. In empirical cases I follow a reactive sequence methodology (Mahoney 2000) commonly used, implicitly or explicitly, in historical sociology (Clemens 2007; Haydu 2010). Reactive sequences are a useful way to study trajectories and paths narratively, i.e. as a sequence of causally linked and temporally ordered events. Key in reactive sequence analysis is to demonstrate the causality of event chains from a key antecedent event to the studied final event (Mahoney 2000). Analysis of reactive sequences is possible with narrative explanation (Abell 2004; Clemens 2007; Haydu 2010), which goes beyond mere chronological documentation of events but orders them into a causally linked storyline. Mahoney (2000) quotes Goldstone’s (1998) case of invention of a first coal-powered steam engine, developed to pump water out of flooded coal mines, as an example of a reactive sequence: “it was just chance that England had been using coal for so many centuries, and now needed a way to pump clear deep mines that held exactly the fuel needed for the clumsy Newcomen pumping machine” (Mahoney 2000: 535). In my cases the final “event” of the event chain is a (quantifiable) sign of decline of the technology in question. A narrative approach is also productive for comparison of historical cases, which is a non-trivial issue for historical studies using variables, but is more amenable in narrative explanations. 6 In the next section I will use three historical cases on which I have elaborated elsewhere (Koretsky et al. 2022; Koretsky and van Lente 2020; Koretsky 2021) to analyse the dynamics of technological decline. 2.4 Comparing cases The first case is a geoengineering technology of cloud seeding. Cloud seeding projects started in the USA (as well as the USSR and the UK) right after World War II. As huge amounts of government and state funding were spent in various 22 Zahar Koretsky projects over 25 years, the technology came under fire after it was used in Vietnam, alongside other disastrous uses of chemicals there. When this was revealed in the press, the negative narratives on cloud seeding (“dangerous technology”, “unethical”, etc.) clashed with the reinforcing ones (“new weapon”, “forest fire relief”, “water supply control for agriculture”, etc.). This clash intensified the discussion on what cloud seeding actually is for, peace or war, with the former winning. A key event for closing the controversy was the signing of an international treaty to stop using cloud seeding for military purposes (the 1977 ENMOD treaty). Cloud seeding funding quickly dropped thereafter. Many federal-funded projects were shut down, but a lot of private and state cloud seeding endeavours continued. There cloud seeding was seen as a fire relief and water supply tool. Much cloud seeding skill and knowledge was preserved in those small-scale and often intermittent projects. In effect, the configuration of “cloud seeding” did not disappear, it just contracted. To grow it, actors would have had to revive and strengthen it. This is exactly what happened in the 1990s as new actors associated a new hypedup meaning of “geoengineering as solution to global warming” to the pre-existing configuration. As a result, by 2010, the seeded area in the US was at least twice as large as during previous peak years. The key events in the history of cloud seeding’s incomplete unravelling were 1) the formulation of a critical narrative by the press and its quick accentuation in Congress and by civil society, who were already agitated by the pacifist and environmentalist movements, 2) the governance decisions to withdraw funding and sign the ENMOD treaty, and 3) the reinvigoration of academic and political debates (e.g. it was routinely brought up in Congress) on global warming, and cloud seeding as a possible response. Cloud seeding is a case of strongly linked configuration, whose unravelling started with internal contestation and weakening in the meanings strand, followed by a similar weakening in the materials strand. The unravelling interrupted when internal contestation within the meanings strand stopped. If no publication about the use of cloud seeding appeared in the press, it would not have or would have much later been picked up by Congress. The timing for the actual discussions was opportune as they coincided with existing environmental and pacifist societal critique. Without (or with a later) awareness in the press and Congress, cloud seeding would have gone unnoticed and would have likely continued to be funded by the government. No international treaty would have been signed or even proposed, or it would have happened much later. The case of cloud seeding’s incomplete and reversed unravelling differs from a more successful unravelling of the incandescent light bulb (ILB) in Europe. The ILB was specifically targeted in the mid-2000s by a coalition of stakeholders which included the industry, the EU and environmentalist groups. Other relevant social groups, the end users at homes and factory floor workers, were not, or only belatedly, consulted. In fact, the latter two bore a high cost for the phase-out, ranging from health effects caused by the replacement lamps, to layoffs, retraining and jobs switches of the workers involved. Although in some countries labour Dynamics of technological decline as socio-material unravelling 23 unions seemed to have cushioned the hit for workers, the numbers of job cuts were still two to three times larger than anticipated by the European Commission. Similar to the cloud seeding case, the unravelling of the ILB started with an attack of a group of actors on the stabilised meanings of the ILB. The view of the ILB as a provider of pleasingly warm, “natural” light was challenged by a critical narrative of energy inefficiency, resonant with a powerful larger discourse of climate change. Within five years, similar to the cloud seeding case, instability in competences and materials followed: massive layoffs, drop of patenting activity, and drop in sales and installation capacity. What is dissimilar to the re-emerged cloud seeding is that the ILB remains in use, care (see van de Leemput and van Lente, 2023) and even production in small pockets of enthusiasts in a downscaled and ever-shrinking state. Thus, the key events in the history of the ILB were 1) the formation of a negative narrative by a powerful social group, 2) their lobbying for anti-ILB (i.e. phase-out) policy, 3) the imitation of the ILB by the LED, and 4) the withdrawal of the ILB to small pockets of production and use. Had the key social group not formulated a negative narrative of the ILB, the European Commission would not have to draft the ban, or at least would do it later. As a consequence, without the sales ban, the ILB would remain on the shelves for much longer. A final case I have studied is the Russian/Soviet original computer series Ural. Ural was one of the most popular computers used in industrial design, medicine, meteorology and banking in the Soviet Union in the 1960s, until it was all but gone by the 1980s (so, well before the dissolution of the country). One of the key events in the unravelling of Ural was Bashir Rameev leaving his post as the lead designer whom other, junior engineers followed. This happened during a notable ministerial rerouting of resources, thanks to a competing priority computer series ES ÈVM. As a result, materials also weakened. Namely, due to a shortage of good quality hardware, Ural designers had to downgrade their initial designs from magnetic disks to inferior punch cards and “punch tapes”. This weakening caused the unravelling. The remaining team resisted the unravelling by attempting to integrate the Ural in ES ÈVM, but were not successful. Because of the difficulties already suffered, there were significant delays with production. Eventually, clients and sponsors withdrew. Subsequently, Ural-related competences slowly dispersed as team members moved on or retired. Without the continuity of the configuration, users rather quickly replaced Urals with ES ÈVMs. The only legacy it had left was the very idea of a Soviet mass computer, adopted by the ES ÈVM. Perhaps, had Rameev remained the leader of the design team, judging from his documented leadership qualities and commitment to the Ural, he could have managed to secure resources for continual development and production of Ural. If so, it would have remained popular and could eventually be integrated into the ES ÈVM. As a consequence, the Ural may have survived until the collapse of the USSR. When comparing these three cases, it is curious to observe that a competing/ alternative configuration was present in the ILB and Ural cases, but absent in the 30 Zahar Koretsky power generation (Turnheim and Geels 2012), and probably many more—perhaps most—technologies. As the cases demonstrate, decline is not a binary switch, but a spectrum between continual performance (i.e., use and/or production) and a lack thereof. In section 2.4 I visualised decline as a radical decrease of intensity of performance across time and space, where unravelling is a vector, or directionality, of technology towards decline. I can now be more specific and demarcate two types of decline indicated by the degree of performance intensity: weak and strong. Weak decline indicates a state where technology is simply not used or manufactured anymore, while all of the strands remain intact. For instance, it may be said that the technology of manually prepared writing ink is in a weak decline because all of the components continue to exist: codified knowledge to manually prepare the ink is noted down in books, the necessary ingredients can be found, and the meaning of “ink” as a coloured liquid material for writing and printing persists. Or in the UK in the 1970s and 1980s (Shove 2012), cycling may be said to have been in weak decline because at least one meaning of the bicycle persisted (transport or exercise), the competences to make and ride bicycles had not disappeared, and bicycles and the lanes were still around. Strong decline indicates the completeness of misalignment in strands and unravelling between them (Figure 2.2), and is thus discontinuous decline. The difference between weak and strong decline amounts to how likely the technology is to return. Figure 2.2 is a rough sketch for illustrative purposes, and “everywhere”, “nowhere”, “always” and “never” should be interpreted not as quantifiable states, FIGURE 2.2 Sketching decline as a relationship between relative geographic and temporal intensity of performance of configuration and types of unravelling Source: Figure modified from Koretsky (2022b) Dynamics of technological decline as socio-material unravelling 31 but as general indicators of high or low geographic or temporal performance in a given geographic and temporal context. In this sense, weak and strong decline, in a given geographic and temporal context, are ideal-types rather than empirical evaluations. Other dimensions, such as the scale of industry and the policy environment, matter, of course; they are not foregrounded in the figure, but are indicated by the different states of the configuration (the three circles). As a next step, the distinction between weak and strong decline allows me to portray unravelling as part of certain (ideal-type) pathways of socio-technical development (Figure 2.3). Note that in the figure the axes are geography and temporality, and the exact shape of the arrow is only an indication. 1. Fall of intensity of performance. The technology is used and/or manufactured less than before, its trajectory is aiming downward. Such is the case of the incandescent light bulb (e.g., in the EU), but also of for example coal-fired power plants (e.g., in the UK) and nuclear power in Germany. FIGURE 2.3 Six non-sequential ideal-type pathways of unravelling overlaid on the axes of geography-temporality (shape of the arrows are for illustration only and will differ per case) Source: Figure modified from Koretsky (2022b) 32 Zahar Koretsky 2. Fall and rise of intensity of performance. Here the configuration remains in the “continual performance” area of Figure 2.2 and no decline takes place. Configurations are in a state of reduced performance, fluctuating between higher and lower performance intensity (in a given geographic and temporal context). One example is cloud seeding which has been seeing a comeback (e.g., in the 2008 and 2022 opening ceremonies of the two Beijing Olympics). 3. Weak decline. Aconfiguration which continues in the direction of decreased performance intensity ends up in a state of “weak” decline in a given geographic and temporal context. Here, the technology is performed so rarely and in such a geographically dispersed way that the associations between the strands are barely there at all. The networks of producers, users and/or maintainers have significantly shrunk. One could say such technologies are on “standby”, although not gone yet. Most technologies, “declined” in the colloquial use of the word, can be placed in the category of “weak” decline, such as nuclear weapon field testing or manually prepared writing ink. 4. Recovery from weak decline. A weakly declined technology may become continuously performed again in a given geographic and temporal context. Such re-emergence could manifest in a reintroduction of the technology, as in the case of vinyl records. 5. Strong decline. When a configuration is performed with less and less intensity (relative to before), this means that one or more of the strands are misaligned. A loss of one or more would lead to a break in the “circuit”. This is a case of a discontinuous or “strong” decline where work will be needed to create or re-create the missing component(s) of the former configuration. Strong decline is more of a hypothetical category because it is hardly possible to ascertain, especially in a large geographic and temporal context, whether materials, meanings or forms of competence are truly gone and will need to be re-created (manufactured, invented, constructed, theorised, learned, etc.). But one may, perhaps, characterise a technology on its way to strong decline when it is found only in museums, e.g., the Ural computer, the 16 mm film (see van de Leemput and van Lente 2023), the linotype machine, kamikaze aircraft, stone arrowheads, medieval and antique construction technologies and other traditional crafts. Strong decline occurs when associations are impossible because slippage, de-anchoring and/or unlearning are too profound (e.g., all materials are destroyed, carriers of competences or knowledge are gone, specific parts of the technology are banned from use or manufacture). After a Dynamics of technological decline as socio-material unravelling 33 strong decline, all that remains are dissociated materials, meaning and competences “debris” in the form of “fossilised” (Shove and Pantzar 2005) rust, hints and secrets. 6. Recovery from strong decline. Even after a strong decline, a technology can hypothetically make a U-turn and re-emerge. From the perspective of the actors involved in the re-creation of materials, meanings and/or forms of competences, this pathway will probably look like the process of invention. Re-emergence in this case occurs when work has been done to (re)construct the strands and re-associate, “re-ravel” them. A weak decline is easier to recover from as to reintroduce a strand (back) to the configuration is less costly. All three cases I have explored above have passed through the decrease of intensity of performance. The decrease has stopped in the case of the light bulb because the ILB has found its niche in industrial applications and horticulture, as well as among people with preferences for the ILB. The light bulb can be categorised as a pathway 1 exemplar. The intensity of performance has actually reversed for cloud seeding, so it is a pathway 2 exemplar. The Ural’s status could be categorised as weak decline because it is not used nor produced as prominently as before. However, the Ural has been delegitimised and for a long time forgotten, and the associated competences are gone with the people who retired and can hardly be found in a codified form. The only intact Ural can be found in a museum and its (cultural) legacy has been largely forgotten. Thus, the Ural may be better categorised as a pathway 5 exemplar, or, bearing in mind the difficulties of ascertaining strong decline, at least on its way to it. 2.6 Conclusions Typically, policy-makers’ focus of nurturing novelties overshadows a focus on initiatives and the ongoing decline of technologies. Such overshadowing might be fuelled by a utilitarian logic of rational progression towards more efficiency. The assumption is that superior technology substitutes the inferior one and the latter then proceeds to fade away, or, as Lindqvist hyperbolised: “[as if] all bicycles [were] thrown in the ditch and all horses shot when the first Model T rolled off the assembly line” (Lindqvist 1994). However, there has been a rise in deliberate decline policies since the late 2010s (Green 2018). Perhaps, actors start recognising that, according to Heyen and colleagues, a “regulated phase-out can offer legal as well as planning certainty for business, workers, consumers and infrastructure planning” (Heyen, Hermwille and Wehnert 2017: 327). My conceptual starting point, drawn from the literature, was to view technology as a socio-technical configuration of entangled socio-material “strands” (materials, meanings and forms of competences). Their dissociation, or unravelling, is proposed as a key mechanism behind decline. Decline, I found, is not, in principle, irreversible, the strands can be re-associated by actors, and this can lead to re-emergence. From this 34 Zahar Koretsky starting point I compared my cases and gained additional insights on how to characterise decline and how to adjust and further develop the conceptual framework. Based on the framework and backed up by an analysis of three cases, I proposed a working approach to study future cases of decline and represent the constituting processes. The proposed unravelling working approach identifies processes within each strand and offers six ideal typical pathways. The proposed working approach will need further fleshing out and adjustment, but already holds, I suggest, four benefits for the field of STS and sustainability transitions. First, in alignment with the symmetry principle, I propose with this working approach to focus more explicitly on a given technology’s “end of life” to counteract the prevalence of studies on emergent technology. Second, I propose that a pragmatic analytical distinction between social and material processes within decline is useful to show their entangled dynamics (cf. Klein and Kleinman 2002; Geels 2007). Third, I propose to focus more on high empirical granularity of analysis and on generation of rich case descriptions that are used towards middle-range theorisation (cf. Frickel and Moore 2006; Wyatt and Balmer 2007). And fourth, I propose to focus more on the analysis of processes outside of market and industry dynamics, since labs, public places, parliaments, newsrooms, kitchens and other sites are equally important (cf. Polanyi 1944; Hughes 1994; Latour 2007; Soete 2011; Upham et al. 2015). Since the presented unravelling approach stems from the STS and transitions literature, it draws, just like this literature, from ecology, complex systems theory, resilience research, ecosystem services, adaptive governance, institutional theory and more. There is room for more connections and adjustments from those literatures for empirical work, theorisation and design of policy recommendations. This chapter is not without limitations. The separation of technology into three “strands” can be questioned, as well as the ontological status of these strands. Specifically, my finding regarding the prominent role of meanings in nearly all presented cases could imply that the separation was invalid in the first place. I would challenge such possible critique with a reminder that a distinction between the three strands is analytical; they were never supposed to be exclusive and nonoverlapping. Of course technology and society are deeply implicated, but the separation is no more than an analytical attempt to navigate the complexities in what causes unravelling and in how it proceeds. Similarly, the somewhat uncompromising collating of different things into the three strands can be questioned. Indeed, this is a radical simplification inherited from the literature on the dynamics of social practice theory. I tried to alleviate this by clearly operationalising what these strands are constituted of and maintaining internal consistency between cases by strictly adhering to this operationalisation. Additional research could no doubt improve the proposed approach on both accounts. Second, the role of actors, structures, infrastructures, levels, knowledge paradigms and ideologies could have been studied more systematically in this work. Future research could go beyond the implicitly flat ontology of the present work and analyse how the different levels of aggregation matter in processes of decline. It Dynamics of technological decline as socio-material unravelling 35 could also be productive to study local situations from the unravelling perspective (e.g., city or regional bans). Transition studies stem from a systems perspective. By definition, commodities, or end products, are not resilient to decline, and are not obdurate; to phase out a commodity is fruitless because it can be, by definition, replaced. This is why destabilisation and phase-out of systems is needed. While I have considered configurations in this chapter, I have not focused on what is typically understood as systems in transitions literature, i.e. configurations that provide societal functions such as mobility, food or energy. The proposed approach would benefit from exploration of such cases, but decline of configurations smaller than systems is (also) an occurring phenomenon worth studying. Although I have not studied these larger socio-technical systems, the presented unravelling perspective might be applicable in studying them too, bringing it closer to the MLP. It is then the responsibility of the analyst to delimit the unit of analysis and maintain it throughout the study. Unruh (2002) writes that discontinuity on a subsystem level (e.g., internal combustion engine and fuel cells) can look like continuity on a system level (transportation). This line of thought may be continued by observing that higher levels of aggregation (e.g., transportation) have more inertia and thus are significantly less susceptible to decline. A specific model of a smartphone is easier to phase out (even though it, being a commodity, will probably be quickly replaced) than the smartphone as a communication technology. Third, the proposed working approach is limited in its ability to prescribe how to trace the “same” technology. A technology changes with time, thus how to know when change stops and unravelling begins? In addressing this limitation, I have, again, followed the rationale in transition studies where the responsibility to define the object of analysis and keep it in focus lies with the analyst (Geels 2011; Stirling 2019). Lastly, although I was motivated to understand decline as a problem of sustainability and decarbonisation within the larger challenge of the climate crisis, the discussed cases only indirectly relate to sustainability. This, however, should not disqualify the observed patterns, and, if anything, expands the programme of understanding and engaging with decline into issues beyond sustainability and decarbonisation. There is no reason why phasing out established systems cannot be equally useful for dealing with ethical questions, for instance responsibility in deliberate decline, nuclear weapons field testing, mass production, surveillance and more (Escobar 2015; Bostrom 2019). Further research could investigate the connection of decline and phase-out with (post-)colonialism, imperialism and other forms of exploitation, both to study the caused decline and the ways to cause decline of these forms. It is a task of future research (see other contributions to this volume) to study more cases of decline of various technologies for the sake of sustainability and topics beyond it. Notes 1 This chapter is based on parts of my PhD dissertation (Koretsky 2022) titled “Unravelling: The Dynamics of Declining Technologies”, where it appears in a more expanded version. This research was supported by Horizon 2020 project FIT4RRI (grant agreement 36 Zahar Koretsky 741477). I thank my supervisors Harro van Lente and Ragna Zeiss, without whose generous support and reviews this work would not have been possible. I also thank Peter Stegmaier and Bruno Turnheim for the review and useful suggestions on the present chapter. 2 Among such prospects, which are reminders of the urgency of the problem, is Paterson’s (2020) bleak forecast of civilizational collapse (absolute declines in human populations, collapse of food production systems, collapse of social institutions) in this century as a result of the climate crisis. 3 I take theory broadly here, including approaches and conceptual frameworks, or as “a statement of relations among concepts within a boundary set of assumptions and constraints” (Bacharach 1989: 496). 4 The distinction between unit of analysis (key focal entity of the study (Sheppard 2020)) and unit of observation (entities to be observed, measured or otherwise studied to learn more about the unit of analysis (Sheppard 2020)) is more typical in mainstream sociology, and here I find it useful for the clarity of methodology. 5 Forms of competence can also be understood from the economics perspective as “intellectual capital”. 6 A theoretical explanation of this point may be found in Haydu (1998) and Mahoney (2000). References Abell, P. (2004) Narrative explanation: An alternative to variable-centered explanation?, Annual Review of Sociology, 30, 287–310. Anderson, K. and Peters, G. (2016) The trouble with negative emissions, Science, 354(6309), 182–183. doi:10.1126/science.aah4567. Ayres, R.U. (1987) The Industry-Technology Life Cycle: An Integrating Meta-Model?Laxenburg. Bacharach, S.B. (1989) Organizational theories: Some criteria for evaluation, The Academy of Management Review. doi:10.2307/258555. Becker, K.L. (2005) Individual and organisational unlearning: Directions for future research, International Journal of Organisational Behaviour, 9(7), 659–670. Bergek, A. et al. (2008) Analyzing the functional dynamics of technological innovation systems: A scheme of analysis, Research Policy, 37(3), 407–429. doi:10.1016/j.respol.2007.12.003. Bijker, W.E. (1995) Of Bicycles, Bakelites, and Bulbs: Toward a Theory of Sociotechnical Change. MIT Press. doi:10.2307/2077312. Bijker, W.E., Hughes, T.P. and Pinch, T. (1987) The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology. MIT Press. Borup, M. et al. (2006) The sociology of expectations in science and technology, Technology Analysis & Strategic Management, 18(3–4), 285–298. doi:10.1080/09537320600777002. Bostrom, N. (2019) The vulnerable world hypothesis. Global Policy. doi:10.1111/1758- 5899.12718. Bourne, M. (2016) Invention and uninvention in nuclear weapons politics. Critical Studies on Security, 4(1), 6–23. doi:10.1080/21624887.2015.1106427. Callon, M. (1984) Some elements of a sociology of translation: Domestication of the scallops and the fishermen of St Brieuc Bay. Sociological Review, 32(S1), 196–233. doi:10.1111/ j.1467-954X.1984.tb00113.x. Callon, M. (1987) Society in the making: The study of technology as a tool for sociological analysis. In Bijker, W.E., Hughes, T.P. and Pinch, T. (eds) The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology: Anniversary Edition. MIT Press. Christensen, C.M. (2006) The ongoing process of building a theory of disruption. Journal of Product Innovation Management, 23, 39–55. doi:10.1111/j.1540-5885.2005.00180.x. Dynamics of technological decline as socio-material unravelling 37 Clemens, E.S. (2007) Toward a historicized sociology: Theorizing events, processes, and emergence. Annual Review of Sociology, 33, 527–549. doi:10.1146/annurev.soc.33.040406.131700. Cohn, M.L. (2016) Convivial decay: Entangled lifetimes in a geriatric infrastructure. In Proceedings of the ACM Conference on Computer Supported Cooperative Work, CSCW. doi:10.1145/2818048.2820077. David, M. and Schulte-Römer, N. (2021) Phasing out and in: System transition through disassociation in the German energy transition – The case of light and coal. Energy Research and Social Science, 80, 102204. doi:10.1016/j.erss.2021.102204. Edensor, T. (2005) The ghosts of industrial ruins: Ordering and disordering memory in excessive space. Environment and Planning D: Society and Space,23(6), 829–849. doi:10.1068/d58j. EEA (2017) Perspectives on transitions to sustainability. EEA. Eisenhardt, K.M. and Graebner, M.E. (2007) Theory building from cases: Opportunities and challenges. Academy of Management Journal, 50(1), 25–32. doi:10.5465/AMJ.2007.24160888. Escobar, A. (2015) Degrowth, postdevelopment, and transitions: A preliminary conversation. Sustainability Science, 10, 451–462. doi:10.1007/s11625-015-0297-5. Frickel, S. and Moore, K. (2006) Prospects and challenges for a new political sociology of science. In The New Political Sociology of Science: Institutions, Networks, and Power. University of Wisconsin Press. Garud, R. and Gehman, J. (2012) Metatheoretical perspectives on sustainability journeys: Evolutionary, relational and durational. Research Policy, 41(6), 980–995. doi:10.1016/j. respol.2011.07.009. Geels, F.W. (2002) Technological transitions as evolutionary reconfiguration processes: A multi-level perspective and a case-study. Research Policy,31(8–9), 1257–1274. doi:10.1016/S0048-7333(02)00062-8. Geels, F.W. (2004) From sectoral systems of innovation to socio-technical systems: Insights about dynamics and change from sociology and institutional theory, Research Policy, 33(6–7), 897–920. doi:10.1016/j.respol.2004.01.015. Geels, F.W. (2007) Feelings of discontent and the promise of middle range theory for STS: Examples from technology dynamics, Science Technology and Human Values, 32(6), 627– 651. doi:10.1177/0162243907303597. Geels, F.W. (2011) The multi-level perspective on sustainability transitions: Responses to seven criticisms. Environmental Innovation and Societal Transitions, 1(1), 24–40. doi:10.1016/ j.eist.2011.02.002. Geels, F.W. (2018) Disruption and low-carbon system transformation: Progress and new challenges in socio-technical transitions research and the multi-level perspective. Energy Research and Social Science, 37, 224–231. doi:10.1016/j.erss.2017.10.010. Geels, F.W. (2020) Micro-foundations of the multi-level perspective on socio-technical transitions: Developing a multi-dimensional model of agency through crossovers between social constructivism, evolutionary economics and neo-institutional theory. Technological Forecasting and Social Change, 152, 119894. doi:10.1016/j.techfore.2019.119894. Geels, F.W. and Kemp, R. (2007) Dynamics in socio-technical systems: Typology of change processes and contrasting case studies. Technology in Society, 29(4), 441–455. doi:10.1016/j. techsoc.2007.08.009. Geels, F.W. and Schot, J. (2007) Typology of sociotechnical transition pathways. Research Policy, 36(3), 399–417. doi:10.1016/j.respol.2007.01.003. Geels, F.W. et al. (2019) Sustainability Transitions: Policy and Practice. Publications Office of the European Union. doi:10.2800/641030. Giddens, A. (1984) The Constitution of Society: Outline of the Theory of Structuration. University of California Press. doi:10.1007/BF01173303. 38 Zahar Koretsky Goldstone, J.A. (1998) The problem of the “early modern” world. Journal of the Economic and Social History of the Orient, 41(3), 249–284. doi:10.1163/1568520981436246. Goulet, F. (2021) Characterizing alignments in socio-technical transitions: Lessons from agricultural bio-inputs in Brazil. Technology in Society, 65, 101580. https://doi.org/10. 1016/j.techsoc.2021.101580. Goulet, F. (2023) The role of alternative technologies in the enactment of (dis)continuities. In Koretsky, Z. et al. (eds) Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation. Routledge. Green, F. (2018) The logic of fossil fuel bans. Nature Climate Change, 8, 449–451. doi:10.1038/s41558-018-0172-3. Grin, J., Rotmans, J. and Schot, J. (2010) Introduction: From persistent problems to system innovations and transitions. In Transitions to Sustainable Development: New Directions in the Study of Long Term Transformative Change. Routledge. doi:10.4324/9780203856598. Hamlin, C. (2007) STS: Where the Marxist critique of capitalist science goes to die? Science as Culture, 16(4), 467–474. doi:10.1080/09505430701706780. Haydu, J. (1998) Making use of the past: Time periods as cases to compare and as sequences of problem solving. American Journal of Sociology, 104(2), 339–371. doi:10.1086/210041. Haydu, J. (2010) Reversals of fortune: Path dependency, problem solving, and temporal cases. Theory and Society, 39(1), 25. doi:10.1007/s11186-009-9098-0. Hekkert, M.P. et al. (2007) Functions of innovation systems: A new approach for analysing technological change. Technological Forecasting and Social Change, 74(4), 413–432. doi:10.1016/j.techfore.2006.03.002. Hess, D.J. and Sovacool, B.K. (2020) Sociotechnical matters: Reviewing and integrating science and technology studies with energy social science. Energy Research and Social Science, 65, 101462. doi:10.1016/j.erss.2020.101462. Heyen, D.A., Hermwille, L. and Wehnert, T. (2017) Out of the comfort zone! Governing the exnovation of unsustainable technologies and practices. GAIA, 26(4), 326–331. doi:10.14512/gaia.26.4.9. Hommels, A. (2005) Unbuilding Cities: Obduracy in Urban Sociotechnical Change. MIT Press. Howe, C. et al. (2016) Paradoxical infrastructures: Ruins, retrofit, and risk. Science Technology and Human Values, 41(3), 547–565. doi:10.1177/0162243915620017. Hughes, T.P. (1983) Networks of Power: Electrification in Western Society, 1880–1930. Johns Hopkins University Press. doi:10.2307/3104214. Hughes, T.P. (1994) Beyond the economics of technology. In Granstrand, O. (ed.) Economics of Technology. Elsevier. Kanger, L., Sovacool, B.K. and Noorkõiv, M. (2020) Six policy intervention points for sustainability transitions: A conceptual framework and a systematic literature review. Research Policy, 49(7), 104072. https://doi.org/10.1016/j.respol.2020.104072. Kierkegaard, M.K. (2014) Lights out for traditional bulbs: Lobbyism and government intervention. International Journal of Private Law, 7(3), 197–270. doi:10.1504/IJPL.2014.062986. Klein, H. and Kleinman, D.L. (2002) The social construction of technology: Structural considerations. Science Technology and Human Values, 27(1), 28–52. doi:10.1177/016224390202700102. Köhler, J. et al. (2017) A Research Agenda for the Sustainability Transitions Research Network. Sustainability Transitions Research Network. doi:10.1016/j.jcp.2005.11.001. Köhler, J. et al. (2019) An agenda for sustainability transitions research: State of the art and future directions. Environmental Innovation and Societal Transitions, 31, 1–32. doi:10.1016/j. eist.2019.01.004. Koretsky, Z. (2021) Phasing out an embedded technology: Insights from banning the incandescent light bulb in Europe. Energy Research and Social Science, 82, 102310. https:// doi.org/10.1016/j.erss.2021.102310. Dynamics of technological decline as socio-material unravelling 39 Koretsky, Z. (2022) Unravelling: The Dynamics of Technological Decline. Maastricht University. Koretsky, Z. (2023) How do technologies die? Studies of decline in literature on technological change. In Goulet, F. and Vinck, D. (eds) Doing Without, Doing With Less: The New Horizons of Innovation. Edward Elgar. Koretsky, Z. and van Lente, H. (2020) Technology phase-out as unravelling of socio-tech- nical configurations: Cloud seeding case. Environmental Innovation and Societal Transitions, 37, 302–317. doi:10.1016/j.eist.2020.10.002. Koretsky, Z., Zeiss, R. and van Lente, H. (2022) Exploring the dynamics of technology phase-outs through the history of a Soviet computer “Ural” (1955–1990). Science, Technology & Human Values. https://doi.org/10.1177/01622439221130139. Latour, B. (1991) Technology is society made durable. The Sociological Review, 38(1), 103– 131. doi:10.1111/j.1467-954x.1990.tb03350.x. Latour, B. (1992) Where are the missing masses? The sociology of a few mundane artifacts. In Bijker, W.E. and Law, J. (eds) Shaping Technology / Building Society. MIT Press. doi:10.2307/2074370. Latour, B. (1999) Pandora’s Hope: Essays on the Reality of Science Studies. Harvard University Press. Latour, B. (2005) Reassembling the Social: An Introduction to Actor-Network Theory. Oxford University Press. Latour, B. (2007) Turning around politics: A note on Gerard de Vries’ paper. Social Studies of Science, 37(5), 811–820. doi:10.1177/0306312707081222. Law, J. (1987) Technology and heterogeneous engineering: The case of Portuguese expansion. In Bijker, W.E., Hughes, T.P. and Pinch, T. (eds) The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology: Anniversary Edition. MIT Press. Levain, A. et al. (2015) Continuous discontinuation – The DDT Ban revisited. In International Sustainability Transitions Conference “Sustainability Transitions and Wider Transformative Change, Historical Roots and Future Pathways”. Levidow, L. and Upham, P. (2017) Socio-technical change linking expectations and representations: Innovating thermal treatment of municipal solid waste. Science and Public Policy, 44(2), 211–224. doi:10.1093/scipol/scw054. Lie, M. and Sørensen, K.H. (1996) Making Technology Our Own? Domesticating Technology into Everyday Life. Scandinavian University Press. Lieu, J. et al. (2019) Transition pathways, risks, and uncertainties. In Hanger-Kopp, S., Lieu, J. and Nikas, A. (eds) Narratives of Low-Carbon Transitions. Routledge. doi:10.4324/ 9780429458781-14. Lindqvist, S. (1994) Changes in the technological landscape: The temporal dimension in the growth and decline of large technological systems. In Granstrand, O. (ed.) Economics of Technology. Elsevier. Loorbach, D., Frantzeskaki, N. and Avelino, F. (2017) Sustainability transitions research: Transforming science and practice for societal change. Annual Review of Environment and Resources, 42, 599–626. doi:10.1146/annurev-environ-102014-021340. Lyle, K. (2021) Interventional STS: A framework for developing workable technologies. Sociological Research Online, 26(2), 410–426. doi:10.1177/1360780420915723. MacKenzie, D. and Spinardi, G. (1995) Tacit knowledge, weapons design, and the uninvention of nuclear weapons. American Journal of Sociology,101(1),44–99. doi:10.1086/230699. Mahoney, J. (2000) Path dependence in historical sociology. Theory and Society, 29(4), 507– 548. McCarraher, E. (2019) The Enchantments of Mammon: How Capitalism Became the Religion of Modernity. Harvard University Press. 46 Bruno Turnheim 3.3 Core notions deployed within transitions studies Destabilisation is related to several notions assuming relative centrality within studies of socio-technical transitions: 1) multi-dimensional sources of stability (inertia and lock-in), 2) multi-dimensional sources of change, 3) incumbents as focal actors, 4) processes, pathways and mechanisms, 5) deliberate or purposeful governance, and 6) vulnerability and politics. 3.3.1 Lock-in as structural and enacted form of stability Destabilisation is essentially a process of departure from or challenge of system stability. Within transitions studies, established socio-technical systems are taken to be relatively stable and coherent. System stability is linked to the notion of lock-in and path dependency, which tend to foreclose opportunities for radical innovation and fundamental reconfigurations in favour of more incremental adjustments to existing socio-technical trajectories. For this reason, transitions remain relatively rare phenomena. On the one hand, structural determinants of lock-in are rather well understood from the perspective of economic rationalities (Seto et al. 2016) and cognitive routines (Nelson and Winter 1982), which include sunk investments (e.g., infrastructure, production facilities), increasing returns from economies of scale (Arthur 1989), industry standards, user externalities, network effects and other positive feedback mechanisms reinforcing asymmetrical advantages accruing from cumulative socio-technical development processes. Non-economic dimensions (e.g., institutional and political) are increasingly being considered (Klitkou et al. 2015; Pierson 2000), but are also admittedly more ambiguous due to the “complexity oftheg oalsofpoliticsaswellasthe looseand diffuse links between actions and outcomes” (Pierson 2000: 260). They include enduring the unusual obduracy of political arrangements, public policies and formal institutions (Pierson 2000), powerful strategic alliances and coalitions with asymmetric access to rule-setting (Roberts et al. 2018), normalised discourses, ideologies and socio-cultural repertoires, and the deep societal embeddedness of user practices and lifestyles. Together, these structural sources of lock-in have the cumulative effect of preventing, limiting or slowing down the development of alternative innovations and socio-technical configurations. Lock-in is an inherently pathdependent process (Arthur 1989; Pierson 2000) resting on self-perpetuating event sequences (Mahoney 2000): past design choices constrain current and future options of system development in ways that favour incremental over radical forms of change and appear as irreversible. But lock-in is neither permanent nor inevitable: system unlocking does and has happened. Understanding and characterising structural lock-in requires 1) longitudinal approaches to how socio-technical configurations have stabilised along particular trajectories to take on specific forms and shapes, and 2) evaluative-descriptive approaches to how stable socio-technical configurations actually are (Klitkou et al. 2015) Destabilisation, decline and phase-out in transitions research 47 and how socio-technical stability may change over time—notably as alternatives are developed or windows of opportunity for change open up. In practice, the situated analysis of lock-in needs to be cognisant of the variety of relevant dimensions and mechanisms (economic and institutional), their relative importance (e.g. structure depth, interlinkages) and their transient nature. On the other hand, socio-technical stability is also enacted by established actors enjoying central or dominant positions within existing systems that tend to reproduce the conditions of their incumbency (Stirling 2019; Turnheim and Sovacool 2020). Firstly, incumbent actors contribute to the continuity of existing systems, structures and practices through tacit and routinised activity, including maintenance, care and repair, continued investment, service improvements, incremental innovation, or the reproduction of underlying (regime) rules. Such continuation, maintenance and improvement activities may confer sustained relevance and legitimacy to the actors involved and contribute to further entrenching related socio-technical practices, arrangements, infrastructures. They are carried out at various levels, from mundane and invisibilised maintenance work to more strategic (and often framed as remarkable) programmes. Secondly, more active forms of stability enactment include strategic activities aimed at defending and protecting current arrangements and advantages (e.g., favourable policy conditions), notably by incumbent actors (see section 3.3.3) in the face of contestation and other challenges to socio-technical stability (e.g., delegitimation). The repertoire of regime resistance and defence strategies can be approached by distinguishing forms of power available to incumbent actors (see Table 3.1). It hence appears relevant, as a preliminary step to better understanding destabilisation processes, to develop the means to evaluate system stability and coherence over time. To do this, it is important to identify markers of system stability and coherence, processes of system stabilisation, and the mechanisms— structural and enacted—that contribute to system lock-in and stability. It then becomes possible to evaluate destabilisation as a process involving the waning of stabilisation mechanisms (erosion of structural forms of stability), difficulties to maintain stability in the face of challenges (failure to actively resist pressures) or intended departure from system preservation objectives and rationales (transformation). The interplay between stabilisation and destabilisation is further discussed in section 3.3.4. 3.3.2 Multiple and mutable sources of change While established socio-technical systems are characterised by lasting stability, they are also exposed to pressures and challenges of varying kinds, which may call into question their normal functioning, expected performance, continued relevance or legitimacy. Such pressures for change may be deflected entirely (e.g., operations resuming after temporary disruption of service), orient incremental improvement and optimisation strategies (e.g., more efficient production modes, substitution of harmful substances), trigger significant adjustments (e.g., new business models, Types of power Related incumbent resistance activities Instrumental Mobilising resources (e.g. finance, capabilities, authority, access to decision-making) in immediate interactions with other actors to achieve their goals and interests Discursive Mobilising authority and legitimacy to shape what issues are being discussed (agenda-setting) and how they are being discussed (issue framing) Material Leveraging technical capabilities and financial resources to promote incremental improvements of existing options over more radical alternatives Institutional Reinforcing prevailing political cultures, ideology and structures (e.g. economic liberalism, technocratic styles) to downplay alternative paths and decision rationales Based on Avelino and Rotmans (2009) and Geels (2014a) TABLE 3.1 Types of power harnessed by incumbent actors for resistive purposes 48 Bruno Turnheim reconfigurations) or eventually make existing systems obsolete. It hence appears important to develop the means to better characterise these sources of pressure, determine and trace their influence over existing systems, namely in order to better qualify what makes for a destabilising source of change (as opposed to other sources of change), and whether specific sources of change may be associated with specific destabilisation patterns. Kinds and sources of (destabilising) change can be related to relevant sociotechnical dimensions within which systems are embedded, the alignment of which confers stability or the de-alignment of which may generate destabilising conditions. The specific dimensions considered vary according to approaches and are largely a matter for analytical choice. These include usual distinctions between techno-economic dimensions (e.g., significant changes in markets, technologies, infrastructures, scientific knowledge), and socio-political dimensions (e.g., significant changes in ideas, policies, politics or cultures) (Geels 2014b; Turnheim and Geels 2013). Socio-technical approaches make a distinction between three interrelated and partly overlapping ontologies (Geels 2004; Geels and Turnheim 2022): that of technical components and systems (artefacts, technical systems, infrastructures), that of institutions (regulations, conventions, cultural values and beliefs, symbolic meanings, and so on), and that of actors and networks (their actions, motives and interests, forms of organisation and interaction). Accordingly, sources of destabilising change include 1) technical dysfunctions, technological discontinuities or performance erosion, 2) social and political mobilisation, delegitimation, the emergence of new rules or the breakdown of existing rules, and 3) challenges by new actor coalitions, the disbanding of existing coalitions or the accumulation of poor strategic choice. Practice theory distinguishes between three kinds of elements that are tied together in stable practices: materials (“things”), meanings (social and symbolic signification) and competences (forms of understanding and practical knowledgeability) Destabilisation, decline and phase-out in transitions research 49 (Shove et al. 2012). So, the decline, erosion or decay of practices can be understood as the unmaking or breaking of ties between such elements. Regardless of the chosen frame of reference, a number of observations can be made. First, challenges to stability can emerge along various dimensions: significant change in one dimension may perturbate the relative stability experienced by established systems. Qualitatively distinguishing different sources of change can underpin the identification of ideal destabilisation processes, including technological disruptions and discontinuities, creative destruction (of economic entities), discursive destabilisation (of prevailing frames), delegitimation (of practices, behaviours or specific actors), de-institutionalisation (of rules and conventions), regulatory challenges and so on. Second, destabilisation is likely to result from a combination of sources of change, so it is more fitting to think in terms of pressure fronts. Third, destabilisation is not merely the result of threats and challenges but is related to the weakening of stability-conferring ties and alignments within socio-technical configurations. Destabilisation results from the combination of intensifying discontinuities (threats and challenges) and weakening continuities (erosion). A related distinction concerns the location and distance of sources of change vis-à-vis the definition of an established system’s boundaries, i.e., whether sources of change are exogenous or endogenous. Exogenous sources of destabilising change typically relate to challenges outside of existing systems and beyond their immediate environment, such as external threats, uninvited challenges or unforeseen discontinuities (e.g., surprises and shocks). Exogenous pressures tend to be less anticipated and not to be the object of dedicated monitoring or intelligence than their endogenous counterparts. “Landscape changes”, such as demographic patterns (e.g., urbanisation), macro-economic trends, geo-political swings, crises and disruptive events (e.g., wars, shocks) are typical exogenous changes. But exogenous changesare notnecessarily macroin scale, as with competition from socio-technical alternatives, social movement contestation, or practice and consumption changes, which can start as relatively isolated and follow more gradual emergence patterns yet come to exert significant pressures for change. Endogenous sources of change, by contrast, are more closely linked to established systems and activities or in their immediate vicinity. They tend to be the object of monitoring (through for example performance indicators); regime actors tend to be more knowledgeable about them andsoina betterpositionto anticipatethemorperceivethemasimmediate threats. Endogenous changes include worsening economic performance (e.g., at product, firm or industry level), declining income, slack (which reduces the ability to manage change) or resources, weakening ties between key sociotechnical components, changes in political support and coalitions, degraded infrastructures (e.g., material, knowledge), and divergence within organisational fields. Furthermore, endogenous sources of change are likely to materialise as conflicts and contradictions between otherwise aligned elements in configurations. However, the distinction between exogenous and endogenous change is not entirely clear cut: it 50 Bruno Turnheim depends on system boundaries, is a matter for interpretation, and exogenous changes can turn into endogenous pressures for change as they become translated into more concrete concerns and pervade socio-technical configurations. Lastly, considerations about sources of change in destabilisation processes raise issues about the explanation of change, and the mechanisms by which sources of change can trigger, reinforce or orient destabilisation processes and patterns (see also section 3.3.4). Since destabilisation rarely results from a single pressure or source of change, its analysis calls for representations of the complex causal chains involved: the temporal interaction of sources of change in the formation of pressure fronts and related destabilisation patterns. 4 With regard to temporality, a variety of possible destabilisation patterns may be identified according to the intensity, speed, scope and sustained nature of pressure fronts. Suarez and Oliva (2005), focusing on changing industry environments, suggest a distinction between regular change, hyperturbulent change, specific shocks, disruptive change, and avalanche change. Stirling (2014), focused on the interpretation of system vulnerability threats, distinguishes between short-term episodic perturbations (shock interpretations) and long-term enduring pressures (stress interpretations). 5 In practice, destabilisation processes are likely to deviate from such ideal-types with interrupted, cycling or reversing patterns. With regard to causation, it is useful to distinguish different levels: proximate causal forces, intermediate causal forces and distal causal forces. The shock/stress distinction illustrates the importance of differentiating causes in practice: the 2008 financial crisis may be interpreted as a shock, the primary cause of which is attributable to the “bursting” of a housing market bubble (subprime mortgages) and of a dependent financial bubble in its wake. But beyond this proximate causation, it is also possible to interpret the crisis as resulting from a long-term stress linked to more remote causes: the accumulation of solvency imbalances (between debt obligations and cash flow), the multiplication of financial innovation, lax monetary policy and an increasing disconnect between the financial system and the real economy. Furthermore, the hysteresis-like pattern of many shocks, while influenced by identifiable feedback loops, is largely emergent and contingent in character. Consequently, it is relevant to approach such processes through the analysis of sequences of events, the identification of critical turning points and possible cascading dynamics of change. So, given that the emergence, evolution and endurance of pressure fronts are an essential source of destabilisation, it is important to develop the means to better characterise, distinguish and evaluate them and their incidence on established systems. 3.3.3 Incumbents as focal actor Destabilisation is intimately tied to the roles, motives and actions of particular social actors, notably those assuming a de facto position of centrality in established sociotechnical configurations and those willing to challenge these positions for various Destabilisation, decline and phase-out in transitions research 51 motives. Many studies adopt a particular focus on “incumbent actors” and their responses to destabilisation challenges (Andersen and Gulbrandsen 2020; Bergek et al. 2013; Berggren et al. 2015; Bohnsack et al. 2020; Dijk et al. 2016; Hess 2019; Hörisch 2018; Isoaho and Markard 2020; Kungl 2015; Lee and Hess 2019; Lockwood et al. 2020, 2019; Mylan et al. 2019; Raven 2006; Smink et al. 2015; Steen and Weaver 2017; Turnheim and Geels 2013; van Mossel et al. 2018), often portrayed as villains to be dethroned or transformed for system change to occur. While the notion of incumbency denotes particular attributes such as a position of centrality, power, mastery over resources in connection to established system, it is important to move beyond simplistic, monolithic and static portrayals of incumbency attributes and the actors that come to incarnate them (Turnheim and Sovacool 2020), to examine a plurality of forms of incumbency at play and the depth of related socio-technical entanglements (Stirling 2019). First, “incumbent actors” is often used as shorthand for large firms or multinational corporations, with little consideration for incumbency, centrality and power in other societal spheres (e.g., policy and politics, civil society, knowledge production). Important questions for destabilisation research remain concerning these other forms of incumbency and how they relate to one another, for instance in incumbent coalitions or constellations perpetuating prevailing paradigms. Second, because incumbency connotes significant vested interests in existing systems, practices and arrangements, it is commonly assumed to exclusively lend to the adoption of resistive postures with respect to radical change, i.e., incumbency as calculated conservatism. In practice, however, the strategic repertoire of incumbents may be broader (e.g., from purely resistive stances to more proactive diversification through various means) and vary significantly from actor to actor. This suggests important questions concerning the range of strategic positions and stances available to and performed by incumbent actors. Third, incumbent strategies are likely to change over time, as the nature of destabilisation contexts changes (e.g., mounting pressures), as specific opportunities arise or as new frames of reference become available to those actors. This raises questions about the changes in incumbent strategies in various contexts and their determinants. Fourth, there is a need for greater clarity as to how the notions of “regime” and “incumbency” are related and intersect. Indeed, certain incumbent actors may “disband” from existing regimes and so contribute to their fragmentation (Steen and Weaver 2017), incumbent actors from neighbouring regimes can significantly contribute to niche construction (Berggren et al. 2015; Späth et al. 2016; Turnheim and Geels 2019), and regime rules are reproduced by all sorts of actors—some of which are not considered as particularly powerful, dominant or central to said regime. So, socio-technical regimes and incumbency only partly overlap. For incumbents, destabilisation ultimately comes with the threat of losing a dominant status, position, sustained relevance or legitimacy (e.g., social licence to 52 Bruno Turnheim operate) but can be met with a broad range of tactical moves and response strategies. Following incumbent actors along destabilisation processes is a relevant analytical entry point, particularly as it enables us to better understand how particular organisations experience, interpret and handle such challenging processes. Correspondingly, incumbent destabilisation processes are being approached from various perspectives deployed at the organisational level, including cognitive and learning approaches, strategic change approaches (e.g., diversification, renewal, ambidexterity), organisational decline, organisational capabilities (Ottosson and Magnusson 2013) or power (Avelino and Rotmans 2009; Geels 2014a). Organisation-centric approaches may further be distinguished according to two dimensions: 1) the depth of challenges and changes (e.g., change in activities, routines, models or core beliefs), and 2) the nature of challenges and changes (e.g., competitive, technological, reputational/legitimacy, socio-political) (see also section 3.3.2). While incumbents are a relevant entry point to study the destabilisation process, other actors warrant attention and dedicated research, including 1) social groups with no or weaker prior links with existing systems and regimes which may actively contribute to destabilisation (e.g., civil society, new entrants organisations, alternative political alliances) notably due to their interest in challenging positions of power for various motives, 2) vulnerable groups with potentially less agency in destabilisation processes such as workers and local communities (Johnstone and Hielscher 2017) and less visible groups directly affected by decline or phase-out interventions (see also section 3.3.6). In other words, there is significant scope for conceptual, analytical and empirical elaborations on agency and power in destabilisation processes. 3.3.4 Destabilisation processes, patterns and mechanisms Destabilisation implies challenges to system stability (see section 3.3.1), is related to different sources of change (see section 3.3.2), and centrally involves actors with strong ties to existing configurations (see section 3.3.3). But at heart, destabilisation is a complex process involving non-linearities, indeterminacies and contingencies. How to make sense, then, of destabilisation as a process, its generative mechanisms, and variety of possible trajectories and outcomes? First, understanding destabilisation as a process requires adopting a mode of enquiry that allows describing and explaining how and under which conditions destabilisation unfolds over time. This implies “understanding both the processes that reproduce durable configurations of social order and those that generate strains or produce events with the capacity to transform social structures” (Clemens 2007: 528), and in what ways the tension between inertial and change forces can evolve. Explaining change in terms of process requires particular attention to (sequences of) events and their enactment: Process research is concerned with understanding how things evolve over time and why they evolve in this way…, and process data therefore consist largely Destabilisation, decline and phase-out in transitions research 53 of stories about what happened and who did what when—that is, events, activities, and choices ordered over time. (Langley 1999: 692) Certain events in destabilisation sequences may be more influential, while others may be particularly significant because of their location within a sequence or because they activate, orient, accelerate or inhibit key processes. Crises may have an accelerating or orienting effect on destabilisation, while the timely development and legitimation of alternatives may contribute to destabilisation by weakening claims about the need to maintain stability at all costs. Further, while the temporal ordering of relevant destabilising events in sequences and their narrative depiction can produce interesting and relevant stories, they also need to lead to generalisable explanations about the phenomenon at hand and hence come into conversation with theoretical arguments: “process theorization needs to go beyond surface description to penetrate the logic behind observed temporal progressions” (Langley 1999: 694)—following inductive or deductive strategies. Process tracing offers a useful approach, empirically focused on individual or small-N cases and analytically oriented towards causal inference through the identification of mechanisms as intervening events (Mahoney 2016). Uses of process tracing vary significantly, and can be mobilised both for theory testing or theory development (George and Bennett 2004). In the case of overly complex and overdetermined processes like destabilisation, recourse to “analytical explanation”, i.e., a “variety of process-tracing converts a historical narrative into an analytical causal explanation couched in explicit theoretical forms” (George and Bennett 2004: 147), appears as a particularly relevant form of process-tracing. The analysis of path dependences, trajectories and branching points along such paths is another useful way forward, provided it does not assume the inevitability of development outcomes (e.g., through consideration of interrupted or resurgent paths) and remains open to counterfactuals. Second, if process tracing can enable the identification of events-as-mechanisms in individual cases, the comparison of multiple cases can improve knowledge claims and conditional generalisations about a given process, particularly if supported by the development of a typology. Typologies seek the identification of relevant causal mechanisms and pathways that influence the outcomes of a phenomenon by specifying its possible variants as ideal-types within a theoretical space. Again, this may follow an inductive logic (i.e., multiple cases allow the identification of different causal pathways), a deductive logic (i.e., theoretical dimensions inform the space of possible pathways), or an abductive logic (i.e. a combination of both in tentative iterations). The elaboration of a destabilisation typology can hence allow specifying the possible causal pathways that destabilisation processes may follow, with particular attention to conditions and contingent mechanisms. Specifying a variety of potential destabilisation pathways can also guide the selection of appropriate case studies, by maximising the variety of observed types or to search for 54 Bruno Turnheim outliers. Importantly, while single-site studies have accumulated evidence of a variety of destabilisation patterns, a useful next step for the research community would be to comparatively locate these within a coherent repertoire of possible mechanisms. Third, the temporality of destabilisation warrants particular attention. Similar to socio-technical transitions, destabilisation typically unfolds over multiple decades (Martínez Arranz 2017), although its duration, speed and pace may vary significantly across cases. In ways analogous to the development of socio-technical systems, destabilisation may be subject to acceleration phases which may wrongly convey an impression of overall speed, so determining the overall duration of destabilisation largely depends on “when one counts” (Sovacool 2016). Periodisation, whether calling upon ideal-typical phases of development or the identification of crucial events, appears as a useful but not unproblematic way forward. Identifying when destabilisation starts and ends is an analytical question. Further, destabilisation, though imprinted in the popular mind by the importance of singular events (e.g., a coal miners’ strike, nuclear disaster, or food crisis), is a cumulative process involving layered temporalities. While tempting to focus only on the most dramatic surface events, those provide “the most distorting and unpredictable lens through which to view reality” (Braudel 1970: 148), and hence require attention to overlapping concomitant temporal processes and nested temporalities (e.g. short term, conjectures, longue durée). Under certain conditions, destabilisation processes might appear as particularly rapid, but in most cases they involve long pre-development phases in which contestation builds up without leading to dramatic struggles, because it is kept at bay or overshadowed by the inertia of incremental system developments. Beneath the highly visible surprises, conflicts and struggles are deep structural tensions, latent disagreements between social groups, and more gradual accumulative processes of problem framing, social mobilisation and political contestation, knowledge and innovation development, the articulation of alternative visions of the future, and so on, which contribute to socio-technical instability. 6 In yet other cases, destabilisation may be entirely uneventful (e.g., gradual erosion or drift) when resulting from a lack of maintenance, under-investment or fatigue. So, destabilisation processes involve a combination of gradual cumulative change and their activation in particular events, which provide focusing devices around which to aprioricen tre empirical efforts, but require investigation of causal chains. Destabilisation is a social process inscribed in multiple temporalities, including long-term social trends, volatile political and public opinion swings, contingent surprises and accidents, medium-term strategic dependencies, abrupt decision reversals and cyclical fluctuations (Braudel 1970; Burke 2005; Sewell Jr 2005), which are combined in particular sequences of events at specific times and places. Understanding destabilisation as a process hence requires explicit analysis of the conditions under which different causal patterns (e.g., distal, intermediate and proximate causes) may accumulate or align. The specification of propositions about such alignments pathways (e.g., their nature, Destabilisation, decline and phase-out in transitions research 55 articulation and timing) is key to crafting destabilisation typologies with significant explanatory power, notably as it allows making sense of varied observable temporal patterns such as gradual erosion (Shove 2012), accelerated destabilisation (Andersen and Gulbrandsen 2020), turning points (Abbott 1997), downward spirals or perfect storms (Hambrick and D’Aveni 1988; Kungl and Geels 2018), or interruptions and reversals (Haydu 2010; Sillak and Kanger 2020). Theoretical models of destabilisation may also attend to the cumulative character of the process by distinguishing different typical stages and outcomes, linked to propositions about expected observations: early stages of destabilisation may be characterised by low or divergent degrees of pressure that can be easily denied or deflected by regime actors (Geels 2014a; Lockwood et al. 2019); moving to later stages of destabilisation may require an accumulation and alignment of pressures as well as alternative path creation (Turnheim and Geels 2013). Such stages are likely to also reflect changes in the strategic positioning of actors involved—destabilisation may require increasing coherence of contestation forces and lead to increasing divergence within existing regimes. However, progression through such stages is neither predictable nor inevitable, and likely prone to pushback from collectives under threat. Destabilisation may hence lead to differentiated outcomes, including partial destabilisation, full decline, reorientation or re-creation (Turnheim and Geels 2012), but also continuity and persistence (Newig et al. 2019; Wells and Nieuwenhuis 2012; Winskel 2018), i.e., when destabilisation pressures are effectively deflected. Fourth, destabilisation is tied to processes of stabilisation in multiple ways. Like stabilisation processes, destabilisation involves ongoing tensions between forces of stability and change, which may ebb and flow, but involves an overall trend away from stability. It can be dialogically related to stabilisation in broader transformation dynamics, particularly in the case of substitution patterns wherein the destabilisation of a given entity enables and requires the stabilisation of another and vice-versa. 7 Destabilisation and stabilisation can also be sequentially related, particularly within the processual logics of punctuated equilibrium (Tushman and Anderson 1986) according to which relatively long periods of stability and incremental change are punctuated by shorter turbulent episodes of radical change in which destabilisation can give way to re-stabilisation, and so on. Given that most transition cases, however, do not follow substitution logics, re-stabilisation is likely to occur at a different structuration level: while the delegitimation of DDT (a powerful and toxic pesticide) led to its subsequent ban (hence fully destabilising the substance’s value and use chain), it however led to new forms of legitimation of pesticides concomitant to the regulation of their use (Joly et al. 2022). This example illustrates the importance of analytical considerations about what is being destabilised, as well as the potential distinction between destabilisation and “exnovation”, 8 notably as they tend to operate at different levels of structure. It also illustrates how destabilisation may lead to the reinforcement of existing structures. Newig et al. (2019), focusing on various types of institutional decline and drawing on Streeck and Thelen (2005a), 62 Bruno Turnheim temporal sequencing (Nilsson and Nykvist 2016). While it is too early to say much about the design of destabilisation policy mixes, it is useful to mobilise Streeck and Thelen’s (2005a) institutional change typology to qualify possible evolution patterns from current arrangements in terms of their structure- and function-preserving qualities in search of evidence of institutional displacement, layering, drift, conversion or exhaustion. Kivimaa and Kern (2016) suggest that while evidence of policy layering is the most likely pattern, what is really needed for destabilisation policy mixes is a displacement and replacement pattern, notably concerning dominant technologies, key actors and rules. Approaching interventions in terms of governance context broadens the perspective to include non-policy interventions, actors and dynamics, as well enabling conditions for destabilisation governance, i.e., the social, scientific, technical and economic dynamics making political projects of destabilisation possible at a given time and over time. What are the conditions under which decisive and sustained deliberate destabilisation governance becomes possible, feasible or desirable? Windows of opportunity for more radical interventions may open up and be seized by change entrepreneurs if problems, proposals and political agendas gain significant traction and become aligned (Kingdon 1984). The dynamic alignment of destabilisation contexts and governance spaces is likely to follow a range of different patterns and is a matter for empirical investigation. For instance, under which conditions do crises become mobilised as opportunities for the deployment of radical phase-out interventions (as opposed to reinforcement of prevailing arrangements)? 3.3.6 Vulnerability and politics This section deals with some of the negative consequences of destabilisation and decline, their linkage to structural problems, and how they might be anticipated and avoided. Indeed, destabilisation raises significant human, social, political and ethical challenges, which may considerably weaken claims of “sustainability 11 ” associated with socio-technical transitions if left unattended, and conversely enhance the emancipatory prospects of transition projects if justice, fairness and redistribution are brought centre stage. Firstly, the destabilisation of socio-technical systems is likely to lead to the decline of certain industries and organisations, the closure of production sites, significant job losses and resulting individual hardships, particularly in communities and regions most dependent on conventional—“unsustainable”—activities. Indeed, by focusing on downfall, closure and abandonment, destabilisation research can usefully support the identification (and anticipation) of potential “losers” of socio-technical transitions. While economic and competitiveness losses in organisations and industries have led to the emergence of “managed decline” arguments and strategies (discussed in sections 3.3.3 and 3.3.5), it is essential to foreground often neglected human and local community perspectives in the process of change 12 —those of “people caught in the cross fire of industrial change” (Cowie and Heathcott 2003: 1). In the context of the British coal phase-out, scholars have pointed towards “the risk of insufficient attention Destabilisation, decline and phase-out in transitions research 63 regarding the broader implications of such discontinuity processes around the impacts on local coal communities and future prospects of the workforce” (Johnstone and Hielscher 2017: 457), by highlighting the uneven impacts that closures have on regional economies, workers and deprived communities. Such concerns are manifest around historically declining activities and livelihoods such as farming or heavy industries, notably given important deindustrialisation and delocalisation trends, for which lived experiences of loss and marginalisation abound and tend to persist well beyond the generation directly exposed to such loss (Strangleman 2017). Further, such experiences take on a new form in the context of environmentally motivated transitions, because of their intentional character—ironically, “deliberate destabilisation” risks not involving much local deliberation 13 —and because of the speed of change called for (Newell and Simms 2020), likely to temporally intensify related challenges. Furthermore, while the material, human and social losses of destabilisation tend to be considered in relation to production systems, sites and jobs, it appears relevant to consider the effects of destabilisation in consumption systems and practices too. For instance, phasing-out certain types of foods deemed unhealthy or environmentally problematic may unfairly affect certain communities over others. Similarly, recent social and political backlashes against the introduction of carbon taxes on petrol (e.g., the Gilets Jaunes movement) illustrate the difficulties of imposing restrictions on daily practices without involving concerned users whose livelihoods may be disproportionately dependent on certain modes of consumption (e.g., car-based mobility in rural or peri-urban communities) in the absence of real alternatives (e.g., affordable and accessible public transport). So, destabilisation raises major social and human concerns, which tend to disproportionately affect communities most dependent on existing means of production and consumption. These communities are hence most exposed to the effects of destabilisation and vulnerable to the potential losses associated with such processes but tend to be less visible and represented in related policy debates. Second, the structural determinants and impacts of destabilisation are tied to particular places, patterns of dependence, vulnerability and opportunity. Concerning social and human dimensions, communities most exposed and vulnerable to the impacts of destabilisation tend to be those with limited resources (income, opportunities, various forms of capital), relatively strong dependence on established systems as means of production and consumption, limited agency or means of representation concerning related strategic decisions, and relatively weak emancipatory prospects (i.e. social groups with little access to alternatives, mobility or relevant infrastructure). So, the human and social impacts of destabilisation are likely to exacerbate existing structural inequalities as well as produce new forms of disenfranchisement. The place-based character of such structural inequalities has been problematised within human geography perspectives (e.g., geographic political economy, evolutionary economic geography), notably in the context of deindustrialisation in highly specialised regions. Deindustrialisation studies tell us two fundamental lessons that may be applicable to destabilisation: rather than an ineluctable process, deindustrialisation is related to 1) a rupture of a long-standing “social contract” between worker unions and 64 Bruno Turnheim management, and 2) a process of capital mobility whereby productive activities are displaced to other localities, activities or forms of investment (Cowie and Heathcott 2003). Likewise, destabilisation can be seen as a process linked to mobility and displacement of local activities (which highlights the local, relative and uneven nature of destabilisation and its linkage to the stabilisation of other forms of activity, rather than viewing it as an irremediable and homogenous process), and a process involving struggles and fundamental changes in long-standing arrangements between relevant social actors (which highlights the political and social choices involved in handling destabilisation). So, the injustices related to destabilisation processes are likely tied to uneven power relations and access to capital between social groups, uneven ties and dependencies to places, and related tensions. Concerning regional economies, sectors and industries, evolutionary economic geography has deployed the notions of regional lock-in (see also section 3.3.1) and relatedness to explain different patterns and outcomes of change in mature industry clusters facing significant challenges (Hassink 2010; Martin 2010). The focus on regional lock-in allows distinguishing adjustment patterns from renewal patterns, and explaining them according to the relative strength of political-institutional resistance to restructuring and to regional economic structure (Hassink 2010). The focus on relatedness (of activities, knowledge and skills in regional economies) enables inferences concerning the adaptiveness or resilience of a regional economy to external shocks: related variety of activities within a regional cluster enables specialisation and growth, but may reduce diversity and adaptability to pressures and shocks (for which unrelated variety may prove more versatile) (Balland et al. 2018; Boschma and Frenken 2011). So, a region’s economic structure may determine its vulnerability to future challenges, regional lock-in may “explain the structural economic problems some old industrial areas face, as well as the related persistence of regional economic inequalities in some industrialized countries” (Hassink 2010: 454), and against observations of path dependence and “path insistence”, it may be possible to engage more constructively with opportunities for regional path creation, path renewal or path development (Hassink et al. 2019). Blažek et al. (2020) have also suggested a need for greater attention to “negative” regional path developments, i.e., those involving the less appealing yet very likely paths of downgrading, contraction or delocalisation. Geographical political economy perspectives highlight how the embedding of regional economies in wider political and economic relations generate interdependences and overflows with negative consequences. Regional destabilisation and decline can result in structural poverty and unemployment for place and people “left behind” (Rodríguez-Pose 2018), notably when alternatives are limited—stranded assets and stranded communities. Even “positive” regional path developments also have their “dark sides” (MacKinnon et al. 2019), indeed dark and bright sides of regional economic development are related in numerous ways (Phelps et al. 2018). New paths can create new forms of exploitation and inequality, because they lead to poor quality jobs (low value, low pay or precarious), dispossession, displacement or inequalities—notably in peripheries and enclaves within Destabilisation, decline and phase-out in transitions research 65 global production networks. Interregional competition can also undermine regional development opportunities in other localities. Third, there is an inherently political dimension to destabilisation because it raises questions about the representation of the most exposed and vulnerable social groups (i.e., those left behind) and because neglecting to do so can lead to knockon effects on socio-political stability, collective hopes and the fabric of society. Policy perspectives on low-carbon transitions tend to emphasise related innovation opportunities and job-creation potential and suggest focusing policy efforts on appropriate jobs and skills development. While sustainability transitions will indeed generate new opportunity pathways for certain sectors, businesses, and forms of employment (e.g., in low-carbon building renovation, renewable energy, or sustainable food production and distribution), they will inevitably also lead to significant job losses in particular sectors and regions (e.g., in extractive, energy-intensive industries or conventional high-input farming). Emerging questions, besides how to re-orient “unsustainable” jobs towards emerging “green” sectors to quantitatively maintain stable employment opportunities, thus concern the scope for 1) avoiding negative and most destructive effects of sustainability transitions on workers and communities historically tied to sectors facing destabilisation, and 2) harnessing transitionsasameanstoenhance the nature, quality, and decency of employment. Such questions have recently been picked up under the framing of “Just Transitions” (Heffron and Mccauley 2018; Jasanoff 2018; Schwanen 2021; UNRISD 2018), to which trade unions and worker organisations have been actively contributing—though such framing still remains peripheral to mainstream policy concerns (Steward 2015). Workers— unionised or not—though perhaps the largest, most vulnerable and exposed social group to the destructive impacts of destabilisation, remain largely invisibilised in related political debates and policy decisions. A worker perspective is however of central importance if we consider that the destabilisation of industries and organisations can lead to a rupture of the social contract between workers and management, that workers have legitimate concerns about how destabilisation is handled, and that worker intelligence about industrial change contributes to creative solutions. Relevant analytical entry points include the role of workers and trade unions 1) in the production of political discourse and framings of destabilisation, 2) in the crafting of strategic responses to destabilisation, and 3) as mediating forces in destabilisation processes. The unequal impact of destabilisation on particular communities is also likely to exacerbate political tensions and fuel political discourses capitalising on forms of injustice. Destabilisation and industrial decline can generate material and symbolic grievances, which populist politics mobilises through place-based narratives (Lizotte 2019). The impact of globalisation, automation, and the weakening of trade union representation of manufacturing workers have contributed to the recent political success of right-wing populist (RWP) parties, with “particular appeal amongst [groups commonly referred to as] the ‘losers of modernisation’ [or] the ‘left behind’” (Lockwood 2018: 718). While the electoral appeal of RWPs is by no 66 Bruno Turnheim means restricted to industrial hinterlands, it is in such places that the more evident link with destabilisation can be observed: The places that don’t matter are becoming tired of being told that they don’t matter and are exercising a subtle revenge. They are voting down or threatening to vote down a system they perceive has quelled their potential and driven them down a road in which the future offers no opportunities, no jobs and no hope (Gros 2016; Rodrik 2017). It is as if the declining agricultural areas and rustbelts the world over have had enough of being patronised and have said, rightly or wrongly, that enough is enough: if we are being told that we no longer matter and that we are going down, the whole ship will sink with us. (Rodríguez-Pose 2018: 199) The rise of global environmental issues as drivers of industry destabilisation have become a new element in this picture, most visible around the climate issue and resulting pressures exerted on high-carbon jobs. Indeed, climate scepticism has become a staple of RWP party discourse, which may be explained through structural arguments related to vulnerability and protection (e.g., hostility to climate policies seen as hitting workers and marginalised communities the hardest) or ideological ones related to anti-establishment and political distrust (e.g., hostility to climate policies seen as the product of a cosmopolitan elite) (Kulin et al. 2021; Lockwood 2018). Regardless, ensuing political discontent and tension around climate policies and “job-killing” arguments are a boon for carbon-intensive industries seeking to lobby more stringent targets and restrictions (Vona 2019). While this arena currently seems largely captured by populist resentment and “places expressing their fear and outrage about potential futures in which they are economically and culturally irrelevant” (Lizotte 2019: 140), there are also reasons to believe that related conflicts and struggle about socio-environmental justice can spur more hopeful forms of political engagement with destabilisation futures. 3.4 Discussion: Conceptual propositions This section builds on the six core notions introduced in section 3.2 and seeks to translate them into conceptual propositions. Together, these provide a synthetic reflection of where research debates stand and some pointers for further conceptual developments. P1: Destabilisation can be seen as a form of challenge, reversal or erosion, of sources of socio-technical stability (lock-ins). Following a distinction between structural and enacted forms of stability, destabilisation involves 1) the breakdown of existing structural patterns, 2) divergence from prevailing action patterns, including a more active mobilisation of resistive power. P2: Destabilisation is likely to result from a combination of multiple and mutable sources of change. Sources of change may vary in terms of Destabilisation, decline and phase-out in transitions research 67 dimensions, intensities, scope or kinds of causation. The intensity of individual and combined sources of change is likely to change over time and, if sustained, may lead to escalating destabilisation pressures. Destabilisation patterns may be distinguished according to the configuration of pressure fronts and their evolution over time. P3: Incumbent actors are central actors of destabilisation with major stakes in and resources for maintaining prevailing configurations. Incumbency is related to important power and resource asymmetries that tend to reinforce stability tacitly (through reproduction of rules, practices and advantages) and actively (strategic moves to resist destabilisation). However, incumbency is also a plural and potentially transient kind of attachment. Other frontline actors of destabilisation processes include those invested in challenging forms of incumbency and vulnerable groups. P4: Destabilisation is a non-linear, indeterminate and contingent process calling for the analysis of causal mechanisms and the conditions of their activation. A process approach to destabilisation should consider how sequences of events and particular conditions can activate causal mechanisms. The development of a destabilisation typology, specifying causal pathways, is a useful way forward, particularly if it is combined with a comparative approach oriented towards the exploration of a variety of pathways empirically. The temporality of destabilisation processes needs to be deconstructed, namely to disentangle punctual crises from more continuous stresses and to make sense of a variety of temporal profiles. Critically exploring the linkages between destabilisation and stabilisation is a useful avenue for theoretical development. P5: The deliberate destabilisation of undesirable socio-technical regimes appears as an emerging, yet politically thorny, horizon for governance.It raises important questions related to 1) limits to agency and coordination vis-à-vis destabilisation processes and their governance; 2) a broad array of (potentially countervailing) motives for destabilisation governance with important implications for what is to be governed and how; 3) a variety of governance instruments extending beyond the conventional remit of innovation policy; 4) the need for thinking through combinations of policy instruments (policy mixes, policy sequencing, policy layering) and the types of governance contexts that may constitute destabilisation as a legitimate object of governance. P6: Destabilisation and phase-out have human, social and political implications, which are exacerbated by uneven patterns of dependence, exposure and vulnerabilities. While such implications are likely to become more visible as destabilisation pressures materialise, they need to be foregrounded and anticipated. Place-based losses are likely to exacerbate existing structural inequalities as well as produce new forms of disenfranchisement. Destabilisation research has much to gain from engaging with place-based accounts of vulnerability, dependence and opportunity. Destabilisation is inherently political, raising issues of political representation and expressions of political discontent. 68 Bruno Turnheim 3.5 Conclusion: Emerging and unanswered problems To conclude, I would like to evoke several unresolved issues and emerging puzzles, as we take a sidestep to engage with the flipsides, the dark sides, the losses and less hopeful aspects of socio-technical transitions. These are what I see as crucial next steps for destabilisation research: my wish list for a collective and distributed research programme. Elaborating typologies of destabilisation pathways. Typologies of transitions pathways have assumed a central role in transitions studies, as a heuristic to make sense of developments in historical cases and to shed light on transitions in the making. Taking destabilisation seriously as a process opens the way for similar typological development. Typologies of destabilisation pathways should be oriented towards the exploration of a variety of possible processes and outcomes, enable pattern recognition in cases, and specify key degenerative mechanisms. This chapter has offered a number of relevant dimensions that such typological work may foreground (see section 3.3.4). Harnessing and expanding empirical variety. Destabilisation research needs to supplement rich single-site case studies with ways to draw comparatively on the growing number of existing cases. To some extent, such work is underway, but it may largely benefit from the specification of destabilisation pathways. Furthermore, it appears important that the comparison of destabilisation patterns extends transversally across sectors, contexts and time periods. A shift towards more systematic comparisons also raises issues about the appropriate unit of analysis and implies trade-offs with deeper engagement with the various scales and sites of destabilisation. So, systematic comparison should not be a substitute for single case studies, but rather go alongside a deep engagement with non-standard cases that can challenge existing concepts and frames. Dealing critically with shocks and temporality. Shocks and crises play a central role in imaginaries, discourse and research related to destabilisation. Destabilisation and breakdown are often ascribed or confined to exceptional circumstances, disruptions and external shocks challenging the otherwise normal operation of systems. Sections 3.3.2 and 3.3.4 have however demonstrated how destabilisation involves multiple and mutable sources of change that may combine in a variety of patterns, ranging from the dramatic to the more gradual or seemingly uneventful. It therefore appears important to situate crises in longer developmental sequences, by attending to what precedes and what may follow crises and the particular circumstances that may lead crises to activate or enable fundamental reconfigurations (crisis/transformation) and those that may lead crises to reinforce prevailing logics and dynamics (crisis/preservation). Section 3.3.5 has further underlined a very fruitful distinction between the discontinuation of governance (in exceptional times) and the governance of discontinuation (Stegmaier et al. 2014), which is an invitation to critically reflect on tensions between preservation and transformation rationales in the face of crises. Dealing with uncertainty, anticipation and imagination. Destabilisation is an inherently uncertain and unpredictable process. That being said, taking the symmetry argument seriously requires engaging with the types of futures that a Destabilisation, decline and phase-out in transitions research 69 destabilisation focus may help envision. An obvious motive is that destabilisation scenarios may be a powerful tool to anticipate, and so possibly avoid, some of the direr consequences of destabilisation evoked in section 3.3.6. For this purpose, destabilisation research needs to engage more with modelling and other types of scenario techniques. Destabilisation futures are not necessarily hopeful and may be inhabited by more monsters than their Promethean counterparts, but there are possibly also joyous and enchanting aspects to future destabilisation, withdrawals and reductions waiting to be uncovered and engaged with. Living with less, without or in the ruins may also have its bright sides—or simply be necessary. Serious engagement with destabilisation imaginaries—dark and bright—appears as a crucial endeavour. Dealing with vulnerability, justice and politics. Destabilisation comes with significant and lasting negative consequences on industries, regions, communities and individuals. To be sure, destabilisation is becoming a very current issue in the context of sustainability transitions, because an innovation-only approach is proving its limits to enable system-wide reconfigurations and because as sustainability innovations overflow out of their niches they are generating new forms of unintended consequences, backlash or resistance from incumbent actors. Transitions studies has always highlighted the role of struggles, but this new phase for transitions (Markard 2018; Turnheim et al. 2018) may be characterised by qualitatively and quantitatively different processes, notably ones in which struggles and politics take centre stage. By proposing a conceptual side-step, destabilisation research affords new ways of thinking about forms of dependence to existing systems, vulnerabilities to change, and the kinds of politics that emerge from situations of impending loss and marginalisation. The just transitions framing is one way forward (Jasanoff 2018), as is thinking through the emancipatory prospects of transitions (Stirling 2015), or thinking more seriously about the social determinants and implicationsofloss(Elliott2018). Dealing with expertise and knowledge. The way it has been framed here, destabilisation research involves an epistemological wager. Shifting the gaze to the existing and its challenges is already delivering significant insights as evidenced by the richness of existing conceptual elaborations (see section 3.2) and promises to bring STS traditions to new ground and debates. As destabilisation research and practice moves forward, it will no doubt spring up new matters of fact (what is destabilisation?) and matters of concern (how should destabilisation be handled?). One yet unaddressed question concerns what this new focus on destabilisation will do to knowledge communities and forms of expertise. Destabilisation bears promises of re-invigorating certain forms of knowledge and expertise around new questions as well as possibly destabilising epistemic communities currently less well equipped to think through its problems. Notes 1 Indeed, in practice, there are very few historical examples of total decline of material or socio-technical systems, which instead tend to lose their centrality and significance as prime engines of socio-technical evolution. 2 Undesirability can be predicated upon various motivations, including environmental, aesthetic, cultural, economic arguments, and always a matter of perspective. 70 Bruno Turnheim 3 It is worth noting that practice bans usually take the form of restrictions on the modalities and locations of allowed use. The notion of perimeter (of use, of ban) hence appears relevant. 4 It is useful to distinguish between different forms of causation. Given the path dependent and contingent nature of socio-technical processes, process theory approaches are likely to be more useful than variance-based approaches, because they attend to the multiple and mutable causes involved in developmental process and event chains rather than attributing a specific outcome or event (e.g. the closure of a firm, the toppling of a political regime) to a set of variables via immutable forms of causation. Consequently, process tracing, event-sequence analysis and narrative analysis are relevant strategies to uncover the processes, contexts and conditions that can explain destabilisation and its outcomes (Smith et al. 2005; Yazar et al. 2020). 5 Stirling’s analysis underscores the importance of interpretive schemes and heuristics as determining the style of action mobilised, with for example shock-interpretations tending towards conservative responses and stress-interpretations towards more radical changes in the underlying conditions. 6 In other words, and following a distinction between causal mechanisms according to the amount of energy (or pressure) they require posited by Bunge (1997), destabilisation may involve a combination of Type I causation (involving energy transfer) and Type II causation (wherein a very small cause may trigger a disproportionate effect). Type II causation hinges upon 1) triggering mechanisms or events, and 2) latent system instability. 7 Such metamorphic analogies can be found in the work of Joseph Schumpeter on the relationship of “gales of creative destruction” wherein industrial mutations involve the breakdown of industrial structures to create new ones—a perspective that has been heavily criticised for its pro-innovation bias (Joly 2019). 8 “Exnovation” (David 2017) or “innovation through withdrawal” (Goulet and Vinck 2017), initially developed as a principle for managing innovation within firms (Kimberly 1981) whereby organisations should devote more attention to divesting themselves of or discarding old innovations to make space for new ones, has become increasingly popular in debates about the deliberate phase-out of undesirable innovations (Davidson 2019; Heyen et al. 2017). While the focus tends to be on the removal of particular technologies or practices, rather than underlying systems and industries, it is pragmatically oriented towards the identification of particularly problematic, vulnerable or changeable elements with the intent to induce wider changes (Newig et al. 2019). 9 Stegmaier et al. (2014), focusing on discontinuation as a purposeful action sui generis, suggest another useful distinction between enacted discontinuation, which consists in setting change in motion, and emergent discontinuation, which consists in seizing prevailing developments of change. 10 Indeed, “abandoned socio-technical systems do not vanish completely and some continued governance effort is necessary long after their exit” (Stegmaier et al. 2014: 121–122). This raises the specificproblem of “governing socio-technical aftercare”,seeking to “control the loose ends of ‘undead’ regime and system parts” following decisions to phase-out, which may include dealing “legally, politically and technically” with remaining stocks and artefacts. 11 Notably by restricting them to environmental benefits, or by losing sight of distributive outcomes. 12 On this particular point, see Andy Stirling’s advocacy of a “worm-eye” view on transition, focused on human experience and practice, as opposed to the more common “eagle-eye” view, focused on top-down management and governance (Stirling 2019). 13 I am indebted to Marc Barbier for intransigently pointing this out. References Abbott, A. (1997) On the concept of turning point. Comparative Social Research, 16, 85–105. Abbott, A. (2004) Methods of Discovery: Heuristics for the Social Sciences. W.W. Norton & Co. Destabilisation, decline and phase-out in transitions research 71 Allan, B.B. (2017) Producing the climate: States, scientists, and the constitution of global governance objects. International Organization, 71, 131–162. https://doi.org/10.1017/ S0020818316000321. Andersen, A.D. and Gulbrandsen, M. (2020) The innovation and industry dynamics of technology phase-out in sustainability transitions: Insights from diversifying petroleum technology suppliers in Norway. Energy Research and Social Science, 64, 101447. https:// doi.org/10.1016/j.erss.2020.101447. Arthur, W.B. (1989) Competing technologies, increasing returns, and lock-in by historical events. The Economic Journal, 99, 116–131. Avelino, F. and Rotmans, J. (2009) Power in transition: An interdisciplinary framework to study power in relation to structural change. European Journal of Social Theory, 12, 543–569. https://doi.org/10.1177/1368431009349830. Balland, P.A., Boschma, R., Crespo, J. and Rigby, D.L. (2018) Smart specialization policy in the European Union: Relatedness, knowledge complexity and regional diversification. Regional Studies. https://doi.org/10.1080/00343404.2018.1437900. Bergek, A., Berggren, C., Magnusson, T. and Hobday, M. (2013) Technological discontinuities and the challenge for incumbent firms: Destruction, disruption or creative accumulation? Research Policy, 42, 1210–1224. Berggren, C., Magnusson, T. and Sushandoyo, D. (2015) Transition pathways revisited: Established firms as multi-level actors in the heavy vehicle industry. Research Policy, 44, 1017–1028. https://doi.org/10.1016/j.respol.2014.11.009. Blažek, J., Kve ˇton ˇ, V., Baumgartinger-Seiringer, S. and Trippl, M. (2020) The dark side of regional industrial path development: towards a typology of trajectories of decline. European Planning Studies, 28, 1455–1473. https://doi.org/10.1080/09654313.2019. 1685466. Bohnsack, R., Kolk, A., Pinkse, J. and Bidmon, C. (2020) Driving the electric bandwagon: The dynamics of incumbents’ sustainable product innovation. Business Strategy and the Environment, 29, 727–743. Borrás, S. and Edler, J. (eds) (2014) The Governance of Socio-Technical Systems: Explaining Change. Edward Elgar Publishing. Boschma, R. and Frenken, K. (2011) Technological relatedness, related variety and economic geography. In Cooke, P., Asheim, B., Boschma, R., Martin, R., Schwartz, D. and Tödtling, F. (eds), Handbook of Regional Innovation and Growth. Edward Elgar. Braudel, F. (1970) History and the social sciences: The long term. Information, 9, 144–174. Bunge, M. (1997) Mechanism and explanation. Philosophy of the Social Sciences, 27, 410–465. Burke, P. (2005) History and Social Theory. Polity Press. Clemens, E.S. (2007) Toward a historicized sociology: Theorizing events, processes, and emergence. Annual Review of Sociology, 33, 527–549. https://doi.org/doi:10.1146/a nnurev.soc.33.040406.131700. Cowie, J. and Heathcott, J. (2003) The meanings of deindustrialization. In Cowie, J. and Heathcott, J. (eds) Beyond the Ruins: The Meanings of Deindustrialization. ILR Press. David, M. (2017) Moving beyond the heuristic of creative destruction: Targeting exnovation with policy mixes for energy transitions. Energy Research and Social Science, 33, 138– 146. https://doi.org/10.1016/j.erss.2017.09.023. David, M. (2018) The role of organized publics in articulating the exnovation of fossil-fuel technologies for intra- and intergenerational energy justice in energy transitions. Applied Energy, 228, 339–350. https://doi.org/10.1016/j.apenergy.2018.06.080. Davidson, D.J. (2019) Exnovating for a renewable energy transition. Nature Energy, 4, 254– 256. https://doi.org/10.1038/s41560-019-0369-3. 4 CONCEPTUAL ASPECTS OF DISCONTINUATION GOVERNANCE An exploration Peter Stegmaier 4.1 Introduction: Focus and phenomenon The fact that socio-technical regimes do not last forever, but decline, is receiving increasing attention in research on the transition of socio-technical systems and innovation dynamics. There is a great need for this attention, because for a long time these research directions have not or only marginally dealt with how existing systems and their supporting regimes (as well as parts of them) destabilise or how innovations break down before they are fully established. The complex interrelationships are not easy to reconstruct. Often, an obvious dimension of this gets lost in the background or is difficult to consider on top of it: the active destabilisation or promotion of destabilisation, the intended exit or termination of support for existing regimes, pursued by a wide variety of actors, alone or together, in the public and political spheres, but just as well in the private and corporate spheres. Not even a reasonably agreed terminology has emerged. 1 Therefore, it must be stated here which terms are mobilised to designate what. “Discontinuation” means the rather actively pursued exit from a socio-technical regime, in contrast to “destabilisation” as rather passively developing (Turnheim 2023). The notion of “decline” itself is used as a very general term to describe the circumstance of progressive destabilisation or discontinuation. The term “phase-out” is employed when it is literally a matter of gradual development. Exit and abandonment are used as relatively non-specific terms to describe the opposite of maintenance, stabilisation, or continuation. Since this paper is about the political, administrative, and managed governance work on an exit from a socio-technical regime, I will mostly speak of “discontinuation” as a more or less purposeful pursuit of a goal. 2 “Discontinuity” then rather refers to the circumstance of a targeted or achieved end of such a regime. Finally, I choose the concept of the socio-technical regime from the multi-level perspective (Geels 2019) as a reference point for the DOI: 10.4324/9781003213642-4 This chapter has been made available under a CC-BY-NC-ND 4.0 license. Conceptual aspects of discontinuation governance 79 governance efforts: as the object-related and theoretical context of choice for the considerations. This chapter is about the purposeful discontinuation of a socio-technical regime, no matter what stage or path of development, maturity, or age. When speaking of socio-technical regimes, I may be talking about narrower or broader domains, the immediate usage environment of a device (e.g., light bulb) as well as the associated value chain (light bulb manufacturing) or embedding in other regimes (household, power supply, lamp design, and many more). I assume that some form of concerted, purposeful action or series of actions, using means of shaping the public or private, will occur if a socio-technical system is to be ended. This can involve large-scale technology as well as small devices, entire infrastructures as well as production lines or product ranges. Discontinuation affects technology as well as the science, politics, economy, everyday practice, or law that supports it. It affects knowledge and ignorance, forgetting and preserving, strategies and routines, individual and collective action. Not everything that happens during discontinuation is induced through governance, but it is a research focus to better understand whether, in which ways, and to which extent governance plays a role. I am in fact interested in configurations that shape the situations bearing the possibility (if recognised and used) to deliberately change/discontinue a socio-technical regime through governance. We still know too little about discontinuation and destabilisation to make a further-reaching exclusionary theory selection. The phenomenon itself cannot be subsumed under existing concepts, patterns, and models—existing research was too often blind to the interplay between discontinuity, socio-technical phenomena, and governance, and therefore the optics for grasping it in an adequate way need to be invented. Since we are talking about a rather new and emergent research field, I suggest, firstly, to use a mix of perspectives and, secondly, to understand these ways of seeing as heuristics—conceptual propositions to see phenomena associated with discontinuation from a governance-related viewpoint. In the following, I will develop arguments about how existing knowledge about discontinuation can be used to argue for certain governance approaches that offer the chance to get to the bottom of discontinuation governance-in-action (section 4.2). I will then introduce a number of relevant existing concepts that can help us to become aware of discontinuation governance, if only as a precursor to casespecific and cross-case concepts that should emerge from the research (section 4.3). I conclude with a series of challenges that discontinuation governance research is currently facing. 4.2 Governance as key focus If we want to put the “governance” of discontinuation on the agenda, we should clarify what can be meant by “governance”. Given the plethora of questions relevant to discontinuation itself, this conceptualisation cannot be exhaustive at this point. Given the newness of the research field, it would also not be appropriate to make a definitional determination. Instead, it is wise to suggest heuristic aspects 80 Peter Stegmaier that might be relevant to consider. Ultimately, it will be a task of any research on discontinuation governance to develop an appropriate concept of governance itself or to specify alternatives such as “policy”, “politics”, “administration”, and so on that fit the respective object of research. However, it would be a pity if no effort was made to specifically shape the concept of governance or if it was only done implicitly. Then this aspect would perhaps be black boxed instead of being made fruitful. The term governance is used to indicate that the notion of hierarchical “steering” or “government action” as the tackling of collective problems falls short of capturing political processes and situations in complex societies, political systems, and states. The questions that regularly arise are: who is doing what, what is it about, in which structures is the action embedded, which structures is it trying to meet or escape and which means and forms of action are used. A general starting point for research with a focus on governance of discontinuation is that by governance we address a process of mutual shaping a political, market, technoscientific, or any other social order. The assumption, then, is that governance-making and governance structures in some political ways have as their aim and purpose the stabilisation, maintenance/repair, and/or destabilisation of a given order. It is political when someone asserts him- or herself alone or with others and achieves a somewhat binding order that is meant to help address collective problems. The various social contexts of these socio-technical orders are also relevant. This applies to any order that a community, an organisation, or a group establishes for itself, more or less on the basis of procedures and rules. This can happen not only by direct control, but also by influencing (or coordinating, negotiating) relationships between more or less independent actors against the background of their resources, arenas, rules, instruments, and in rather heterarchical or rather hierarchical contexts. Does governance mean a more tentative or directive approach (Kuhlmann et al. 2019)? We would want to see which relationships characterise a governance and how, which orientations and motivations are at work and/or expressed, and how (far) individual actions are coordinated. This we would need to understand also in temporal contexts, so what was before the current situation and in which direction it moves. Here it is helpful to look at the development in its patterns and manifestations as well as individual situations in the course of this development. The not-so-frequent nor well-known discontinuation governance especially must be searched for how an opportunity actually arises that discontinuation becomes central. Which failed attempts, abortions, preparations, changes of plans, power structures and discourse struggles first opened up the chance to distance oneself from something familiar, “normal” and accepted for a long time? Do stable orders break down from delegitimisation, disinterest, non-action or destruction, withdrawal or redefinition of knowledge and norms, perception, or interpretations? Does a regime destabilise from dysfunctionality, erosion, exhaustion, weakness, or fragile statics, from the inability to be adapted or updated (Joly et al. 2022)? In addition to process and situation orientation, the analysis of problem orientation is also cultivated (Kohoutek et al. 2013). It examines how discontinuation Conceptual aspects of discontinuation governance 81 governance takes place (as process), in which constellation it asserts itself and ends a socio-technical regime (situation), and finally along which problem perceptions and agendas it finds its authoritative orientation. It doesn’tsuffice only to mention many aspects that need to be looked at more closely typically, but also to point out symmetries. This means that when the focus is on governance-making, one should not ignore the structural circumstances; or, conversely, that one should not overlook the specific action of influence or triggering action in the face of all the structures and abstract developments; and that one should put events and justifications, knowledge- and value-related perceptions, explanations, and justifications in relation to each other. One can read volumes about the usual policies and governance arrangements, routines, and improvisations. But what is discontinuation—how do you do it, why, what for, what triggers it, hinders it, or ends it? How far is discontinuation presenting itself as a governance innovation? In order to study how discontinuation is actively addressed by governance, it needs to be seen how much public and private or mixed elements are de-aligned; how much corporate (private) governance and public policy (alone or together) are defining the subject of discontinuation (e.g., of actors in a household or in public transport). It must be clarified for each individual case in which planetary, global, regional, national to local interconnections discontinuation takes place. How are specific and meta-governance frameworks related, are their developments interdependent or do they contradict each other? Is a meta-governance framework built up first before a specific regime is brought down? 4.2.1 Discontinuation as a social construct When governance tackles something, we observe that it is typically because those involved have agreed that there is a problem that needs to be addressed (Colebatch 2006; Colebatch et al. 2010; Dewey 1927). The fact that there is a problem for a collective, what it is for them, how to recognise it, what it has to do with other or previous problems, how one usually reacts to it, and so on are all questions of shared perception, interpretation, and negotiation of what to do with it (Kingdon 1984; Hoppe 2010) in the light of norms and theories, patterns of consciousness and frames as our (shared) interpretations of how situations are organised. Conflicts also arise as to what one sees, whose view is correct or whether/how one usually reacts to it with governance. Normative questions also arise whether one should see something one way or another, what one should do with it, what is the right (factually appropriate) or best (politically, morally, ethically, etc.) reaction to it. 3 If one considers these processes of the social production of occasions to engage in governance, it becomes clear why one would do well to explore how active or non-active discontinuation occurs, how it comes about that discontinuation governance does not always receive the same attention, although it may already be a means of choice again and again on a smaller scale. It is therefore also worthwhile to see how discontinuation governance can experience an upswing when, for 82 Peter Stegmaier example, as is currently the case, the most diverse upheavals are being discussed and discontinuation is seen as a way to go, as in energy issues, technology development, environmental protection, or industrial policy (or for other areas, see, e.g., Sato 2002; Bauer et al. 2012; Princen et al. 2015). If both the way of acting, which can be understood as discontinuation governance, and the occasion to do so, must be continuously produced and maintained, i.e., are not determined as innate or natural law conditions, then we can see them as social constructs. Then we can say discontinuation governance is communicatively constructed, be it at the beginning or as it is sustained during social negotiations (Eberle 2019: 265). Negotiation is thereby understood as a generic category referring to the negotiated order as a result of all kinds of interactions, not only as one form of shaping social order among others (cf. Strauss 1978, 1993: 57). Following Berger and Luckmann (1966), we can see discontinuation governance and its result, discontinuity, as the co-construction of a discontinuous reality: as objective facticity and subjective meaning. This means asking how it is possible that human activity should produce or discontinue a world of things and vice versa: the co-construction of continuous or discontinuous reality (ibid.: 27). This means that something is encountered as a social matter that cannot easily be ignored or escaped, and that this has meaning for an actor. For instance, in 2009, at the start of the phaseout, the incandescent light bulb (ILB) was an object that followed norms that were increasingly perceived as delegitimate, an object that industry wanted to get rid of, and the ILB users knew how to interpret this situation, e.g., by stockpiling the ILB for use beyond the end of production because users liked the light quality, or, conversely, they shared the perception that ILBs needed to be replaced for energysaving reasons and hurried to exchange the old ones for the new ones. At the same time, it is clear that precisely discontinuation and destabilisation entail a change and thus a partial dissolution of an existing social order (in our case better: socio-technical order) that has hitherto been preserved grosso modo. The fluidity of order, order as something mobile, plastic, turns out to be sometimes lossy while analysing and governing discontinuation and destabilisation: not everything remains forever. What just belonged can lose legitimacy or even become obsolete. 4.2.2 The agency–structure nexus There are many reasons to assume that discontinuity is not purely a question of active governance action. Socio-technical regimes are too large and complex to always be brought down with simple actions. Where active action is taken with this aim in mind, passive tendencies may also come into play, intended developments may go hand in hand with unintended ones, and planned developments may go hand in hand with unplanned ones. Active attempts may fail, take a long time with an uncertain outcome, or have a partial rather than comprehensive impact. Existing destabilising tendencies may first be discovered, waited for, and eventually exploited, but not triggered completely on their own. Before active action can take effect, a great deal of preparatory work may have to be done, such as discovering Conceptual aspects of discontinuation governance 83 and picking up on existing trends, exploiting other processes for discontinuation, and laying a variety of foundations so that discontinuation can become conceivable, communicable, acceptable, realisable, and justifiable. Times and windows of opportunity to become active can be wrongly chosen or less sustainable than thought. At the same time, however, one cannot claim that discontinuities occur without any intervention or active influence, impetus, or intention. Socio-technical regimes may be doomed to decline because they are not maintained and developed, or even neglected, or because they overlook problems that over time build up to existential threats. Complex systemic relationships can lead to discontinuation and destabilisation just as well as any other condition. Destabilisation may not lead to the end (Turnheim 2023); this may need active influence (cf. Le Quéré et al. 2019). Decline could ultimately be accepted—otherwise one could try to resist it and keep something alive, pour financial resources into it and declare war on its ending. From this point of view, the failure to intervene and resist—the failure to act, the veto by doing nothing—can also be an active measure 4 (cf. Weber 1978; Streeck and Thelen 2005). Turnheim (2023: 56) calls this “inaction” a “particular form of intervention: neglectful or self-consuming governance”. But one can see how decisions are made that lead to discontinuity, at least in the areas subject to this decision-making power. When NASA pulled out of the space shuttle programme, realistically speaking, it will not have been the decision of the NASA leadership alone, nor will it have been made entirely of its own free will. But a formal act of termination certainly took place. When Siemens sold its nuclear division, this did not mean that nuclear technology ended everywhere, but that it ended at Siemens—together with the management and knowledge structures, the personnel and the value chain that was once built up and maintained. Thus, it seems advisable to search for those combinations of active action and existing developments, of new and existing processes and efforts that make the governance of discontinuity tangible. It will be a matter of identifying situational as well as processual, structural, and mobile aspects of the governance of discontinuity. In practice, there are hardly any cases in which the public or private can manage entirely on their own. On the contrary, we regularly see that these areas are somehow intertwined—for example, when Siemens sold its nuclear division as a company, this was a corporate matter closely linked to one country’s national nuclear policy; when the government closes nuclear power plants, this affects the energy industry, among many other things; households must be prepared for changes in electricity quality and prices, and must figure out whether or not to support a more sustainable energy policy. The underlying problem is, in brief, that people act, resulting in structures that in turn influence action (Berger and Luckmann 1966; Giddens 1984). It is therefore important to consider how discontinuation governance as action reacts to existing structures or questions old ones, dissolves them, or even creates new ones. Action and structures intertwine. Structures of discontinuing action are individual pieces or types of actions and actors, concatenations, clusters, habitualisations, 84 Peter Stegmaier institutionalisations, socialisations, knowledge structures and distributions, and so on. If a discontinuation has already taken place, any further discontinuation already refers to prior knowledge, patterns of action, even instruments and other institutionalised (repertoires of) routines equipped with rules and knowledge of rules (Cairney 2012; Colebatch 2006). Or there is recourse to modes of governance that do not come directly from discontinuation; work is done analogously to other governance. Combinations of exceptional and analogue modes of governance can be found. From a governance analysis perspective, it is also important to remember that not all problems are perceived or addressed by decision-makers, partly because it would be too burdensome to address them or break them down into manageable parts. Thus, starting to exit can seem like an oversized hurdle because it has never been done before (if that is indeed the case) or could trigger consequences (sideeffects, collateral damages) that encompass more than just the discontinuation issue. 4.2.3 The things–meanings nexus Discontinuation governance targets a spectrum from perceptions, frames, knowledge structures on the one hand, to built environment, architecture, and infrastructure on the other hand. What has been introduced as a social construction of discontinuity is realised both in the breaking up of thinking patterns and dismantling of what has been built in stone and steel. It is about breaking down previously valid knowledge and old certainties as well as dismantling production plants, buildings, and the structural realisation of infrastructure in hardware and information technology structures such as software. What has been said so far about governance can also be applied to the terms policy and politics (or whatever distinctions the respective contextual language offers). It is not so important which of the terms one uses, but for what and how. Much of what is said about policy today comes very close to the concept of governance, especially the departure from top-down and governmental focus (Colebatch 2009). One can also see the work on problem conceptions, policy alternatives and decision-making politics as dimensions of governance (Kingdon 1984). Governance is also realised in material and digital structures. 4.3 Existing conceptualisations There are several concepts that have been around for a while to help understand decline in governance, policy, or administration studies. These are the two prominently represented terms—“termination” and “dismantling”—but alongside them the lesser-noted should also be mentioned. In the following, some typical perspectives in research on science, technology, innovation, and policy will be identified. It will be discussed which aspects might be helpful for our collection of building blocks for a conceptualisation of discontinuation governance. It starts with concepts from the field of policy and governance studies and ends with references from science and technology studies. In between, I highlight the emergence of discontinuation Conceptual aspects of discontinuation governance 85 aspects in socio-technical transition studies as well as in studies on socio-technical economic, innovation, and more recent discontinuation governance studies. 4.3.1 Discontinuation concepts in studies of governance, policy, administration Although there is some attention for questions of discontinuation, work on “ter mination”, “dismantling”, “reversal”, “removal”,or “retrenchment” never reall became a core topic in policy studies research and textbooks. The abandonment o dismantling of socio-technical systems, deeply embedded in society and the econ omy, has not been studied yet from that angle. 5 Our focus is somewhat differen from these existing literatures, however, for two main reasons: first, the focus isn’ton particular actor or social group (firm, industry, or an entire policy branch such a welfare), but on technologies and socio-technical systems and the regimes that carr them that involve and interact with many actors. More importantly, we are intereste in a particular kind of discontinuation, which is driven or motivated by the desire on th part of relevant governance actors to address one or more perceived particular social issue (e.g., protection of the EU lighting industries and climate change, if you think of th incandescent light bulb phase-out within the EU ecodesign framework; safety an environmental effects with regards to the ban of DDT by the Stockholm Convention - - y r - t a s y d e s e d ; safety and risk issues in the case of the German nuclear phase-out). The perception of adverse effects of a specific socio-technical system is at the core of the mobilisation process. It very often requires production of knowledge. Second, our emphasis is on understanding by whom and how “issues” are being articulated, framed, and put on political agendas in related actor arenas—in other words, how “problem”-driven political agendas manage to question and challenge incumbent governance arrangements within and around a socio-technical system, and possibly may result in deliberate discontinuation governance efforts,which maybemoreorlesssuccessfulovertime. Policy termination In the policy studies literature, there is some attention for the governance of discontinuation of particular policies and programs, often referring to the term “policy termination” (cf. Bauer 2009). Policy termination may either result from a changed formulation or perception of a policy problem, or from a changed formulation or perception of a policy solution (van de Graaf and Hoppe 1996: 211–227; Stegmaier et al. 2014: 114; on perception change regarding dam removal, see Clark 2009: 403). This can easily be translated into “governance termination” as the termination of governance that contributes to maintaining a socio-technical system, if we want to opt for the broader notion of “governance” that encompasses policy as one stream next to problem recognition and politics, as well as the focus on the broader collectivity of those involved (cf. Kingdon 1984). Studies of policy termination offer some addition to our insights and hints for further research. Bardach conceives of the politics of policy termination “as a special 86 Peter Stegmaier case of the policy adoption process” which “is exceedingly difficult” (Bardach 1976: 123; cf. Turnhout 2009). He compares different cases of programmes, policies, and organisational entities with less or more technoscientific orientation, including the US participation in the Vietnam War. He distinguishes the “explosion type endings” occurring “through a change in policy effected by a single authoritative decision” from the “very long whimper” type coming “not from a single policy-level decision but from a long-term decline in the resources by which a given policy is sustained” (ibid.: 125), the latter gradual development is also called “decrementalism” (Lambright and Sapolsky 1976). Termination meets “coalitions of proponents and opponents” in contest over terminating a policy (Bardach 1976: 126), dividing the first into “Oppositionists”, “Economizers”, and “Reformers”, the second described as the guardians of the status quo in administration, political system, and organised interest. A first obstacle for easily deciding to pursue termination seems to be that policies are often built on a nimbus of having an infinite future (ibid.: 128). Second, opponents to termination, Bardach suggests, mobilise useful interrelationships between political and moral order (ibid.: 127), as expressed “in the general moral repugnance… towards the deliberate disruption of arrangements which people have learned to rely on for a significant portion of their livelihoods or careers” (ibid.: 128). Third, Bardach suggests policymakers are “reluctant to admit—or seem to admit—past mistakes” (ibid.: 129). Fourth, pro-termination coalitions seem to hesitate to damage an existing regime (ibid.: 129). Finally, some regimes seem to reward novelty and innovation rather than termination—whereby the latter is apparently not so easily regarded as the actual innovative approach. Termination is facilitated, Bardach suggests, by “change in administration”, “[d]elegitimation of the ideological matrix in which the policy is embedded”, “turbulences” which shake optimistic expectations, “cushioning the blow” in case of policy failure or heavy crisis, and the experience that strategic “designing policies for eventual termination would facilitate their transformation or even their complete destruction when the time was ripe” (Bardach 1976: 130). Frantz (2002) reconstructed for a case how a government has used resources and political argument to make a policy termination successful: how goals were articulated, how problems were defined, and how a variety of solutions to the problem were used. Frantz discovered that it is not, as often assumed, the government and administration that hold the key to attempts at termination at all. She shows how much “the skilful use of political argument and resources empowered the opponents of closure in their battle against the outwardly more powerful forces of the government… how the government succeeded only when it made better use of its political resources” (Frantz 2002: 25). In another study, Frantz (1997) carved out how the high costs of policy termination come about. She starts from the observation that often “policy terminations are promoted as cost saving moves when, in fact, there are often considerable short term costs” (Frantz 1997: 2097), both monetary and psychological (cf. Behn 1978; deLeon 1978). Frantz, based on a study of termination in a public health system, categorises costs to prevent damage to community, constituents, and staff (Frantz Conceptual aspects of discontinuation governance 87 1997: 2111). In a strongly economically dominated governance-making environment, it would be extremely important for discontinuation governance research to know 1) what costs are incurred, 2) how they are calculated, and 3) how they compare with (cost) alternatives to discontinuation solutions (cf. also Joly et al. 2022: 45). Choices, framings (cf. Liersch and Stegmaier 2022), concerns, justice perceptions (cf. Johnstone and Hilescher 2017), resistance against discontinuation (cf. deLeon 1978), among others, could thus be better contextualised. In this context, it is also relevant to see whether and which measures are taken to cushion the impact of discontinuation (cf. Hogwood and Peters 1982), which is probably one key aspect of a special governance to discontinue another governance as opposed to governance that is dedicated to coordinating and deciding about the process. Again for health policy, Sato (2002) and Sato and Narita (2003) reconstructed the abolition of the leprosy prevention law in Japan in the 1990s through leadership. Sato (2002) introduced the figure of the “skillful terminator” who set the issue on the agenda, negotiated, achieved consensus, and got the thus delegitimised policy abolished. This I would call a “discontinuation entrepreneur” (a variant of a policy entrepreneur or institutional entrepreneur, who seeks policy initiative and change as an antidote to persistence or inertia in a field under pressure to change). Interestingly, the role of science remained instrumental and limited (Sato and Frantz 2005). How expertise is used to initiate or justify discontinuation is a relevant question for research. Policy dismantling Policy dismantling stands for “a distinctive form of policy change, which involves the cutting, reduction, diminution or complete removal of existing policies” (Bauer and Knill 2012: 31) with the two dimensions of policy intensity and policy density. Bauer et al. (2012), on social and environmental policy for instance, emphasise the quantifiable shrinking of spending for a policy branch, which as a consequence of this leads to, for example, a dry out of welfare state measures or a weakening of air pollution control. “Policy dismantling” is usually used as an opposite term to “policy expansion” (cf. Knill et al. 2018), whereby Gravey and Jordan (2019) also distinguish “limited expansion” and “limited dismantling”. Both authors also present a dismantling strategy typology, which helps to better grasp active dismantling action: “no dismantling decision”, “passive dismantling strategies”, “active dismantling decision”, and “active dismantling strategies” (cf. Bauer and Knill 2014). This can be a key typology also for discontinuation governance-in-action. There is also attention in policy literature for the dismantling of public administration institutions when a state lacks a system of integrative political institutions (Vogelsang-Coombs and Keller 2013). Gravey and Jordan (2019) are shifting the dismantling focus from a mostly national to an international level. Policy dismantling touches upon technoscientific questions when addressing environmental (Lenschow et al. 2020) and energy policy (Bürgin 2018; Barnett et al. 2020; Prontera 2021). In addition to this advanced concept of dismantling, we often find 94 Peter Stegmaier climate change and justice (Ayling and Gunningham 2017: 136). Neville observes that because of the perception that fossil fuels destabilise the climate, divestment considers “fossil fuel markets as risky for investors” (2020: 4). 8w?>According to the economic logic of divestment, the external costs are internalised, which leads to investors moving to other areas and recognising that sustainable management leads to savings and new opportunities (cf. Halstead et al. 2019). Consequently, Neville sees the resulting dismantling of the fossil fuels regime, which in turn may lead to a transformation of related regimes, the economy, and society. Neville suggests that “[d]ivestment efforts delegitimize the fossil fuel sector, countering its political and cultural power” (2020: 4; cf. Bergman 2018). The questions that follow for further discontinuation governance research are to what extent these causal attributions are empirically confirmed in each case, which corporate and public governance frameworks are created or used for this, and how the systemic destabilisation and the actively pursued regime discontinuation are weighted. From a financial political viewpoint, it should also be asked how and how far the readjustment of “political economic practices of financialization, capitalization, and assetization” (Neville 2020: 8) through divestment leads to again growth or degrowth, and whether and how this changes these practices. What is divestive discontinuation in practice and as a regime arrangement? Other research considers the role of international institutions, such as OECD, OPEC, WTO, UNFCCC, and IEA, in governing the transition away from fossil fuels (van Asselt 2014), non-state governance in climate policy (Ayling and Gunningham 2017), and open confrontation at extraction sites over exiting from investments or keeping them going (Curran 2020; Liersch and Stegmaier 2022). Exit In management and organisation theory, there is literature on the decline of particular organisations (Cameron et al. 1987; Caves et al. 1984; Mone et al. 1998; Lamberg and Pajunen 2005), which tends to focus on unsuccessful adaptation (e.g., because of an inability to overcome inertia or “wrong” strategies) to changing environments. Ghezzi (2013) distinguishes enterprise-driven (and firm-specific) from environment-driven discontinuities. In much of this literature, the decline of an organisation is seen as a long-term process that goes through different phases. Decline is seen not only as an inevitable result of external impulses that can hardly be ignored, “but also as an endogenous and strategic reaction to performance problems” (Turnheim and Geels 2012: 37). Weitzel and Jonsson (1989) suggest distinguishing the five stages of “blinded”, “inaction”, “faulty action”, “crisis”,and “dissolution”. They argue that the lack of an appropriate response leads to the decline of organisations. The literature on organisational decline also emphasises that there is not one decline path, but several possible pathways depending on the nature and speed of external pressures and endogenous responses. Turnheim and Geels (2012, 2013) have actively drawn on this literature and theories to make sense of destabilisation and decline at the level of industries and inter-organisational fields. Conceptual aspects of discontinuation governance 95 Recent management studies tend to view exit as a natural part of a company’s overall strategy (cf. Cefis and Marsili 2007; Graebner and Eisenhardt 2004; Villalonga and McGahan 2005; Mohamad et al. 2015). From this perspective, exit is not necessarily synonymous with failure since the decision to exit can also mean an increase in efficiency or the realisation of profits (Cefis and Marsili 2007: 1; Dupleix and D’Annunzio 2018). Obsolescence Obsolescence reminds us of the fact that discontinuity is caught between planned and unplanned, desired and undesired. In the area of consumer products, the loss of functionality, which leads to the replacement or disposal of the product, has a positive character from the point of view of manufacturers and sellers, who have an interest in products not lasting forever and not becoming hits on the second-hand market, so that new products need to be purchased. Studies could show to what extent obsolescence is an expression of the throwaway culture 6 that is obsessed with disposing of products (while they could just as well be maintained, repaired, mothballed, used as a source of spare parts, museumised, etc.). In other areas, the end of usability is more associated with a loss of investment and market share. Legislators and consumers may each have opposing stances on the industry view. Of course, there are consumers who like to replace old clothes with new and fashionable ones; old electronic items with new ones with updated features. 7 But at the same time, consumers may be more interested in ensuring that what they buy doesn’t break down too quickly, because otherwise they lose more money than they have or want to spend (Kuppelwieser et al. 2019). The legislator stands between them if, on the one hand, they want to promote production and trade (and skim off taxes), i.e., they do not necessarily promote the longevity of products, but, on the other hand, the legislator may also want to limit the unbridled throwaway mentality on sustainability considerations in some areas (Ober et al. 2017). However, this attribution of obsolescence is not only based on how far it is planned, but also on replaceability with new things and dysfunctionality in the face of changed circumstances (Mellal 2020). It is carried out for products as well as for production plants and processes, organisations, and bodies of knowledge (Warmington 1974). Mellal also points to a form of obsolescence that is still often overlooked, namely the situation where an alternative would be available but it does not move forward even though it could. The question is why we continue with the old; what political conditions are set so that the new alternative is not chosen and the old technology is not terminated; and how much discontinuation governance knows about such hidden alternatives based on existing but unused obsolescence. How to deal with the fact that knowledge becomes obsolete (Margulles and Raia 1967) when a socio-technical regime comes to an end (and something else takes its place) is also a question discussed as a problem of obsolescence, but only very little for discontinuation and destabilisation. Linked to this is the even more complicated situation when this occurs, but remnants of the old regime remain and 96 Peter Stegmaier need to be taken care of. In this case, the old knowledge is largely fading away, but it still must be preserved in relevant parts in order to be available for special uses, maintenance obligations, waste products, or late effects. We will find here a variety of interconnections between business and public, economic, and political occasions and practices, structures and instruments that aim at or at least entail discontinuation. Thus, this area calls for in-depth observation, analysis, and discussion of where 1) obsolescence is employed or abandoned, desired, or not tolerated, 2) how this discontinuation is engineered or subverted in practical terms, and 3) how this is related to political and normative ideas about how it could be justified to let some things break faster or less quickly, to accelerate or stop the decline. 4.3.4 Discontinuation in studies of innovation regimes Studies of innovation regimes and sustainability policies have also discovered discontinuation. They treat it not only with exit and destabilisation terms, but also with the distinctive concepts of “exnovation” and “outnovation”. The difference between the notions of ex- and outnovation seems to be that while exnovation is used to address the life-cycle end as the end of an innovation, outnovation addresses the letting go of a part of a larger socio-technical system in a move that can in itself be quite innovative in doing things differently than before. Exnovation occurs rather passively, outnovation deliberately. Exnovation The notion of “exnovation” was originally used in terms of innovation through external impulses, 8 as a quasi-branded consultancy term. Recently, it has also been used in environmental sociology to contrast the introduction of the new against the discontinuation of the old (Kropp 2014; David 2014; Gross and Mautz 2015), and the social and governance nature of the process is emphasised that abolishes technologies thereby making space for new ones (Paech 2013; Sveiby et al. 2012). David and Gross (2019) emphasise that “exnovation” addresses the “‘natural’ flipside of innovation”. From the perspective of the firm, “exnovation” was apparently first introduced by Kimberly (1981) who described innovation as a series of processes which in combination define an “innovation life-cycle”. The final process of the model is exnovation, where the organisation must discard existing practice associated with a previously implemented innovation, thereby allowing the adoption of a new innovation, where the life cycle starts again. (Patterson et al. 2009: 27) A fruitful aspect implied is that innovation can become obsolete. The question for governance and management is whether there is enough capacity and capability to end Conceptual aspects of discontinuation governance 97 one innovation or maintenance path and turn to another. Problematic with the “exnovation” notion is that authors often jump directly from in- to exnovation, for instance, when explaining that “exnovation implies active rejectionof aninnovationthathas been invested in previously” (Gross and Mautz 2015: 3; Kimberly 1981: 91), no mention of the fact that what was an innovation might have been the norm for quite some time before it becomes questioned to such an extent that its discontinuation might become visible on the governance horizon. Furthermore, although seen as a governance task, there is hardly any specification of an “exnovation governance” (be it public or corporate governance). Outnovation The notion of “outnovation” has been used by Levain et al. (2015) to consider discontinuation as a specific type of innovation, in terms of an “inverse innovation” or “innovation through withdrawal” as suggested by Goulet and Vinck (2012). Outnovation means a change, which proceeds through the withdrawal or subtraction of a part—or the whole—of a socio-technical regime (Pellissier 2021: 172). Pellissier suggests the notion of “outnovation” is an act of innovation policy that removes a substance or product from the regime, as was the case for DDT: innovation by substitution (although in other cases it could also lead to other pathways). Levain et al. (2015) and especially Pellissier emphasise the active part of registering and banning a pesticide—as public action. Pellissier shows that the use of pesticides is not free but must be carried out in accordance with rules; that registration does not distinguish between products that are considered toxic and those that are not, but selects those whose use seems to be compatible with the registration rules. A tension can thus be seen within the registration system, which, on the one hand, allows products under certain conditions, but, on the other hand, can prohibit them (Pellissier 2021: 172). In this way, Pellissier also raises questions about the continuity of regimes. She explains that bans on pesticides enrich the set of instruments of controlled use, while the regime does not change radically, but may even be reinforced (ibid: 195– 196). However, Pellissier also points out that the prohibition of a substance or product for which no alternatives by substitution exist, or do not yet exist, could lead to a different situation. Prohibition without substitution challenges the dialectic between innovation and prohibition (ibid.: 196). It will be an important task of research on decline, discontinuation, and destabilisation to find contrasting cases and to identify the constellations and processes and then learn to distinguish between them. In all these empirical questions, the concept of “outnovation” is synonymous with that of discontinuation. Authors who use the notion of discontinuation, however, do not emphasise that it is about the countermovement to an innovation, but refer to any termination, whether of an incomplete, recently completed, tentatively established, or firmly established regime or one that is increasingly perceived as weak and undesirable. 98 Peter Stegmaier 4.3.5 Discontinuation governance studies: Pathways and configurations As result of their comparative studies of discontinuation governance, Stegmaier et al. (2023) view “discontinuation” as a property of a trajectory 9 in which the constituting relations become misaligned to such an extent that the trajectory’s distinctive character is lost, as one possible result of various permutations of distributed agency, emergence, contingency, or deliberate governance. 10 Discontinuation governance entails two complementary aspects, in which groups of governance actors undertake towards and eventually within a window of opportunity a spectrum of actions (including use of policy instruments) intended directly to effect both the discontinuation of a trajectory itself (governance of discontinuation) and of governance practices that help stabilise it (discontinuation of governance). The governance of discontinuation is here understood as governing the discontinuation of a particular governance problem as the result of a changed framing (formulation, perception) of a socio-technical regime (cf. van der Graaf and Hoppe 1996). The discontinuation of governance practices, in turn, is seen as the discontinuing of a particular way of solving a policy or a governance problem as the result of a changed framing (formulation, perception) of a problem or solution. Their research found that a trajectory can become the addressee of discontinuing (besides and most in contrast to “building” and/or “maintaining”) governance. Since the empirical reality of governance targets is complex, often composed of nested entities, and the targeted trajectories are rather “moving targets”,the “targeted trajectory” (e.g., DDT, incandescent light bulbs, or nuclear energy production) is distinguished from the “wider trajectory” (pesticides, electric lighting/energy efficiency for energy-using products, and the nuclear industrial-technical-military complex). When a phase-out or ban occurs, some characteristic patterns can be seen: specific configurations that evolve over time (pathways). Both are important: the configuration as a particular structural state, and the emergent character of it when on the way. While these patterns could also be used to describe dynamics that emerge, they are intended here as markers of strategic directions of governance action: 1) how these modes of influence on a socio-technical target object are used in a targeted way, or 2) how a bundle of measures leads recognisably towards this direction. In contrast, the phase model of destabilisation enactment (Turnheim and Geels 2012), for example, can be understood as a reactive or response scheme, in which it is largely a matter of reacting appropriately and counteracting destabilisation, but if this does not help and the de-alignment trend intensifies, at least still accompanying or managing the decline. The discontinuation governance pathways, also in their specific sequence in each case, follow more the logic of a proactive scheme which is based on the assumption that discontinuation is intended and should therefore be fostered. When a “window of opportunity” opens, different patterns can be observed: � Ending pathway: Two forms of governance-induced ending, often associated with the notions of “phase-out” and “ban”, which are the incremental or the abrupt discontinuation of a trajectory. Conceptual aspects of discontinuation governance 99 � Weakening pathway: Three preparatory modes of discontinuation—control (producing intelligence, limiting use by critical observation/social control effect), restriction (scope of usage), and reduction (scope of production)—which together can lead to final discontinuation in terms of ban or phase-out, but which can also “only” lead to a development that persists, but in a somewhat constrained, retrenched, or limited way. These pathways cover the core of each discontinuation governance phenomenon: the way to ultimate closure, at once or in steps. Another set of pathways captures discontinuity in relation to continuity after a window of opportunity has opened and has been used or not used: � Expiration pathway: Two possible progressions, either what is discontinued gets replaced (substituted)—here, the relation of the discontinued trajectory to the one that takes its place is part of the discontinuation governance; or abandonment occurs, which means that discontinuation governance is independent of anything that could take the place of the old and that the gap it leaves is not filled directly. � Continuance pathway: When no discontinuation occurs at all, in cases in which efforts of discontinuation governance are induced and a destabilisation of the trajectory could be observed, but this doesn’t lead to discontinuation (not yet, not within foreseeable time, or due to flaws in the process or advocacy coalitions breaking apart before success). These types describe rough alternatives. In the individual cases and with even more detailed historical considerations, it will probably be found that some trajectories do not completely disappear for a long time. What is left is often a remnant of usage and knowledge, infrastructure, and function, for a transitional period, until all its functions are replaced by new ones (think of plastic production, which cannot be done without petroleum products, if the goal is fossil-free production). At other times, it may be a matter of decision, for example if DDT can still be used for very limited purposes or if lightbulbs are intended for special uses. In short, it seems as if almost nothing disappears completely at first (cf. Stegmaier 2023). 4.3.6 How (far) STS grasps discontinuation and its governance Abandonment of technologies and socio-technical systems occurs not infrequently. However, similar to governance, administration, and policy studies, science and technology studies is not known for a broad consideration of discontinuation and destabilisation. The emergence of novel technoscience in the form of scientific, technological, medical, engineering knowledge, artefacts, and practices is usually prioritised over all forms of decline. However, it is Kuhn (1962) who had already posed the key question in his studies on scientific revolutions: 100 Peter Stegmaier What is the process by which a new candidate for paradigm replaces its predecessor? Any new interpretation of nature, whether a discovery or a theory, emerges first in the mind of one or a few individuals… How are they able to, what must they do, to convert the entire profession or the relevant professional subgroups to their way of seeing science and the world? What causes the group to abandon one tradition of normal research in favor of another? (Kuhn 1962: 144) The empirical cases are legion. However, it is crucial to see how socio-technical systems, technological regimes, or technologies are (or have been) disappearing or are being ended. Research on discontinuation and destabilisation has not yet made a real breakthrough. The authors of the studies of the Ural computer (Koretsky, Zeiss and van Lente 2022), unravelling (Koretsky 2023) and 16 mm film (van de Leemput and van Lente 2023) set out to change this. And they have precursors who have developed important clues, as we see below. Trajectories of erosion, decay, and fossilisation Shove and Walker conceptualise the “trajectories of erosion, decay, and fossilisation” (Shove and Walker 2007: 767) as parallel movements to innovation. This prominent and extensive focus on decline in STS turns out to be at the same time a particularly problematic way of prioritising progress. It overlooks the fact that, firstly, innovations can also be aborted; secondly, that sometimes innovations are deliberately intended to function as bridging technologies (halogen lamps as a temporary replacement for conventional incandescent lamps, as well as energy-saving and LED lamps as long as they were not yet fully functional; hybrid drives as an intermediate step to electric cars; gas as an intermediate step to all-round energy supply with renewables while nuclear and coal are discontinued); and thirdly, that innovations of governance are in some cases newly developed in order to be able to undermine or bring down a strongly established existing regime. Moreover, Shove and Walker limit themselves to such downward trajectories as evolutionary rather than managed or controlled processes; to decline as the result of an inherently uncontrollable process (Pantzar and Shove 2010: 459). Shove portrays emergence and novelty as bright, whereas the unmaking, erosion, and decay as dark—the “shadowy side of innovation” (2012). Shove reproduces the “conceptual and empirical emphasis on novelty” (2012: 364) that she criticises. Unviability However, there are in STS a few important studies addressing the issue of ending directly. Latour presents the drama of the end of the automated train system Aramis in France, which he compares to the Concorde supersonic aeroplane in sideremarks (Latour 2002/1992). In each of the cases, their respective technologicalpolitical dimensions are profoundly linked. One of the socio-technical regimes was Conceptual aspects of discontinuation governance 101 abandoned before its introduction, the other only after three decades of operation, during which it was always controversial. The termination of political support for the Aramis project led to the death of the socio-technical system as a whole. When it had almost matured sufficiently, the political will to use it was still lacking—and it never came. To fully mature (for example, the complex software) would still have required a not easily determinable time and an enormous additional budget. Some parts that were too complicated had already been simplified during the development process, while others, such as the control software, were not. Latour makes an important observation here: that when projects become too complex, individual parts can be left out without terminating everything (discontinuation of unviable components). If this is not done in sufficient numbers, 11 the whole system could become unwieldy, too complex, and possibly be up for disposal. What was not managed was both technological sophistication and the excessive expectations of the stakeholders—a tension between technology maturation and project management that is difficult to resolve and points to the close link between governance and technology: the specific concoction of the socio-technical regime. In the end, it’s hard to decide whether Aramis was feasible, either technologically or regarding its governance. It is a question of perspective whether one sees Aramis as a collapse of innovation or the end of a short lifetime. The fundamental patterns do not seem very different. Concorde, the civil supersonic aircraft, was a technically functioning system, albeit with quirks, which was only terminated after a fatal accident at Paris Charles de Gaulle airport after 27 years of operation. This is remarkable because the end had already been threatened during its development: the costs of production and operation were high, acceptance of the enormous noise pollution among the population in the neighbourhoods of airports was low, and potential customers did not want to take the acceptance risk (Japan cancelled an order because of this). In 1973, the US even banned civil supersonic flight. Concorde was a politically charged intergovernmental project of the French and British governments, from which the British partners did not withdraw only because they did not want to incur the wrath of the French over the necessary breach of contract. With Concorde, then, we register a tense mixture of political will (France) and political unwillingness (UK), technological maturity (basic functionality) and technological inadequacy (noise). Moreover, the political-regulatory framework changed due to bans on civil supersonic flight. It can be said that Concorde was kept alive artificially for some time (with great effort, with many popular rejections), while Aramis did not even see the light of day, but was stopped before birth. De-inscription and non-use For cases in which technologies are already in use, script analysis may offer another lead, e.g., when Akrich and Latour (1992) refer to “de-inscription” as one form of rejecting or renegotiating what is prescribed by designers’ programmes for users. While designers assume projected users with typical forms of use, interests, 102 Peter Stegmaier competences, and taste, and inscribe this vision like a “programme” in an artefact (Latour and Woolgar 1979; Akrich 1992), actual users may not comply with the programme of action and the delegation of roles. Users may then elaborate an alternative “anti-programme” and use it to de-inscribe the original programme thereby reshaping the artefact. The artefact as such will not (necessarily) be abandoned. Through the concept of the gender script, it is proposed to capture all the work that involves not only the inscription but also the de-inscription of representations of masculinity and femininity in technical artefacts (Oudshoorn and Pinch 2003). In many studies there are some elements embedded that relate to governance or policy. Discontinuation comes in when users refuse to use or change the inscribed, intended programmes for use. In that respect, the slowly growing scholarship on non-use and no-longer-use after long use, or after test use before full adoption, always has the potential to be re-read for a discontinuation governance focus. Here, non-users are seen as active agents in the destabilisation of technologies during or after introduction, before or after longer use (cf. Oudshoorn and Pinch 2003; Wyatt 2003; Melby and Toussaint 2016; Weiner and Will 2016). Conceptual discontinuity The ANT approach with its symmetrisation of human and non-human entities among many other categories as a methodological programme and objective breaks with conventional presuppositions about the separation of nature, technology, and society. Instead, hybrids such as the ozone hole are considered, composed of CFCs, refrigerators, industrialists, politicians, chemists, and meteorologists (Latour 1993). In this way, ANT not only lays the groundwork for methodological and theoretical rethinking, but also political potential and, importantly in this paper, the inclusion of politics and governance, administration, and regulation in STS analyses. The diagnosed crisis of the moderns also becomes the level of consideration for the crisis of large-scale, unsustainable business-as- usual modernisation and other challenges that politics and governance are already forced to consider together in practice. In the same way, feminist STS have offered challenges to rediscover the customary order of gender relations and to re-inform and re-address sociopolitical issues of notions of gender, assigned function and performance in science and technology. Breaking down gender stereotypes goes hand in hand with breaking down the practices and rules that assign an unequal role to female members of society. Gender research suggests discontinuation needs. Later, new biotechnological and medical STS research showed how nature can become both a construction and demolition site for outdated concepts of the body, gender, science models and the boundaries of technology (vis-à-vis bodies and brains). Postcolonial research showed how American-European or Western-Northern science and technology, in conjunction with politics and policy, have provincialised the horizon and what might be de-universalised in response. With this rough overview, I only want to hint at the great and broad potential of STS to address in much greater depth numerous questions relevant to the political shaping of society, also from the perspective of the leveraged, deconstructed, delegitimised, and discarded thought models and socio-technical regimes. Conceptual aspects of discontinuation governance 103 Doing with less or without Goulet and Vinck note that responses to the problems facing modern industrialised and technologised societies involve reducing or even abandoning certain technologies, substances, or other artefacts essential to ways of life and production; and that this is even happening at an increased rate. They suggest a critique of the sociology of translation in that it tends to overemphasise the making of new connections and the breaking of existing associations (attachments). They focus on mechanisms of dissociation and detachment. To speak of mechanisms of dissociation and detachment suggests a view that sees not whole substantial terminations or exits, but the making of cut-offs, disconnections, reductions, dissolutions, and distances from each other of elements that have hitherto been seen as belonging together. In this way, parts remain, while others no longer have a function, or at best do so in a different place and in a different way. The aim is to better understand the detachment processes at work in most innovations (Goulet and Vinck 2012). In doing so, Goulet and Vinck refer to an entire discipline of innovation studies from Schumpeter to organisational studies, transition studies, regime dynamics to Callon, which consider innovation without a detailed look at withdrawal and destruction. The focus is on complementing innovation research with the withdrawal perspective. Key forms found are banning through delegitimation and prohibition, as well as disintermediation as “removal of an actor or an intermediary object… considered harmful” (Goulet and Vinck 2017: 108). Maintenance and care as forms of doing less with new things are only recently receiving more attention (see van de Leemput and van Lente 2023), yet almost never in terms of the ending of care-taking or maintenance endeavours, such as the category of “aftercare” for the care-taking of some remaining components of a socio-technical regime until they are also finished, as introduced by Stegmaier et al. (2014) to address typical policymaking practices in the discontinuation context. 4.4 Challenges for research about discontinuation governance From the preceding considerations, we identify the following key challenges for describing and operating discontinuation governance. Why both? Because, on the one hand, we still have a lot to learn about discontinuation work and circumstances and, on the other hand, because practitioners are also regularly faced with the question of identifying challenges and opportunities for action. 4.4.1 Extinction: Discontinuation in studies of bio-physical / sociotechnical-ecological regimes A largely neglected field in studies of decline, discontinuation, and destabilisation is the reduction of biodiversity and extinction of species as a side effect or declared goal of management and governance actions. This is also about the overexploitation of nature through clearing, settlement, and reclamation, through the destruction of 110 Peter Stegmaier Geels, F.W. and Raven, R. (2006) Non-linearity and expectations in niche-development trajectories: Ups and downs in Dutch biogas development (1973–2003). Technology Analysis and Strategic Management, 18(3–4), 375–392. Geels, F.W. and Schot, J. (2007) Typology of socio-technical transition pathways. Research Policy, 36(3), 399–417. Geels, F.W. and Turnheim, B. (2022) The Great Reconfiguration: A Socio-Technical Analysis of Low- Carbon Transitions in UK Electricity, Heat, and Mobility Systems. Cambridge University Press. Ghezzi, A. (2013) Revisiting business strategy under discontinuity. Management Decision,51 (7), 1326–1358. Giddens, A. (1984) The Constitution of Society: Outline of the Theory of Structuration. Polity. Girod, B., et al. (2014) Climate policy through changing consumption choices: Options and obstacles for reducing greenhouse gas emissions. Global Environmental Change, 25, 5–15. Goulet, F. and Vinck, D. (2012) L’innovation par retrait: Contribution à une sociologie du détachement. Revue Française de Sociologie, 2(532), 195–224. Goulet, F. and Vinck, D. (2017) Moving towards innovation through withdrawal: The neglect of destruction. In Godin, B. and Vinck, D. (eds), Critical Studies of Innovation: Alternative Approaches to the Pro-Innovation Bias. Edward Elgar Publishing. Graebner, M.E. and Eisenhardt, K.M. (2004) The seller’s side of the story: Acquisition as courtship and governance as syndicate in entrepreneurial firms. Administrative Science Quarterly, 49(3), 366–403. Gravey, V. and Jordan, A.J. (2019) Policy dismantling at EU level: Reaching the limits of “an ever-closer ecological union”? Public Administration, 98(2), 349–362. Gross, M. and Mautz, R. (2015) Renewable Energies. Routledge. Haeuber, R.A. and Mitchener, W.K. (1998) Policy implications of recent natural and managed floods. BioScience, 48(9), 765–772. Hall, P.A. (1993) Policy paradigms, social learning and the state: The case of economic policymaking in Britain. Comparative Politics, 25(3): 275–296. Halstead, M., et al. (2019) The importance of fossil fuel divestment and competitive procurement for financing Europe’s energy transition. Journal of Sustainable Finance & Investment, 9(4), 349–355. Harkins, P. and Prior, N. (2021) (Dis)locating democratization: Music technologies in practice. Popular Music and Society, 45(1), 84–103. Hillman, A.L. (1982) Declining industries and political-support protectionist motives. The American Economic Review, 72(5), 1180–1187. Hogwood, B.W. and Peters, B.G. (1982) The dynamics of policy change: Policy succession. Policy Sciences, 14(3), 225–245. Hoppe, R. (2010) The Governance of Problems: Puzzling, Powering and Participation. Policy Press. Hyysalo, S., Jensen, T.E. and Oudshoorn, N. (eds) (2016) The New Production of Users: Changing Innovation Collectives and Involvement Strategies. Routledge. IRENA and ILO (2021) Renewable Energy and Jobs – Annual Review 2021. International Renewable Energy Agency, International Labour Organization. Johnson, S.E. and Graber, B.E. (2002) Enlisting the social sciences in decisions about dam removal. BioScience, 52(8), 731–738. Johnston, P. and Hielscher, S. (2017) Phasing out coal, sustaining coal communities? Living with technological decline in sustainability pathways. The Extractive Industries and Society, 4, 457–461. Joly, P.-B., et al. (2022) Gouverner l’arrêt des grands systèmes sociotechniques. In Goulet, F. and Vinck, D. (eds) Faire sans, Faire avec Moins: Les Nouveaux Horizons de L’innovation. Presses des Mines. Conceptual aspects of discontinuation governance 111 Kemp, R. and Loorbach, D. (2006) Transition management: A reflexive governance approach. In Voss, J.-P., Bauknecht, D. and Kemp, R. (eds) Reflexive Governance for Sustainable Development. Elgar. Kern, F. and Howlett, M. (2009) Implementing transition management as policy reforms: A case study of the Dutch energy sector. Policy Sciences, 42, 391. Kern, F. and Smith, A. (2008) Restructuring energy systems for sustainability? Energy transition policy in the Netherlands. Energy Policy, 36, 4093–4103. Kimberly, J.R. (1981) Managerial innovation. In Nystrom, P.C. and Starbuck, W.H. (eds) Handbook of Organizational Design, Vol. 1. Oxford University Press. Kingdon, J.W. (1984) Agendas, Alternatives, and Public Policies. Little, Brown. Knill, C., Steinebach, Y. and Fernández-i-Marín, X. (2018) Hypocrisy as a crisis response? Assessing changes in talk, decisions, and actions of the European Commission in EU environmental policy. Public Administration, 98(2), 363–377. Köhler, J., Geels, F.W., Kern, F., Markard, J., Onsongo, E., Wieczorek, A., Alkemade, F., Avelino, F., Bergek, A., Boons, F. and Fünfschilling, L. (2019) An agenda for sustainability transitions research: State of the art and future directions. Environmental Innovation and Societal Transitions, 31, 1–32. Kohoutek, J., et al. (2013) Conceptualizing policy work as activity and field of research. Central European Journal of Public Policy, 7(1), 28–59. Koretsky, Z. (2023) Dynamics of technological decline as socio-material unravelling. In Koretsky, Z. et al. (eds) Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation. Routledge. Koretsky, Z., van Lente, H., Turnheim, P. and Stegmaier, P. (2023) Introduction: The relevance of technologies in decline. In Koretsky, Z. et al. (eds) Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation. Routledge. Koretsky, Z., Zeiss, R. and Van Lente, H. (2022) Exploring the dynamics of technology phase-outs through the history of a Soviet computer “Ural” (1955–1990). Science, Technology & Human Values. https://doi.org/10.1177/01622439221130139. Kropp, C. (2014) Exnovation – Nachhaltige Innovationen als Prozesse der Abschaffung. Unpublished manuscript. Kuhlmann, S., et al. (2019) The tentative governance of emerging science and technology – A conceptual introduction. Research Policy, 5, 1091–1097. Kuhn, T.S. (1962) The Structure of Scientific Revolutions. University of Chicago Press. Kuppelwieser, V.G., et al. (2019) Consumer responses to planned obsolescence. Journal of Retailing and Consumer Services, 47, 157–165. Lamberg, J.-A. and Pajunen, K. (2005) Beyond the metaphor: The morphology of organizational decline and turnaround. Human Relations, 58(8), 947–980. Lambright, W.H. and Sapolsky, H.M. (1976) Terminating federal research and development programs. Policy Sciences, 7(2), 199–213. Latour, B. (1993) We Have Never Been Modern. Harvard University Press. Latour, B. (1996) On actor-network theory: A few clarifications. Soziale Welt,47(4), 369–381. Latour, B. (2002) Aramis, or the Love of Technology. Harvard University Press. Latour, B. and Woolgar, S. (1979) Laboratory Life: The Social Construction of ScientificFacts.Sage. Le Quéré, C., et al. (2019) Drivers of declining CO2 emissions in 18 developed economies. Nature Climate Change, 9, 213–217. Lejon, A.G.C., Renöfält, B.M. and Nilsson, C. (2009) Conflicts associated with dam removal in Sweden. Ecology and Society, 14(2), 4–22. Lenschow, A., Burns, C. and Zito, A. (2020) Dismantling, disintegration or continuing stealthy integration in European Union environmental policy? Public Administration, 98(2), 340–348. 112 Peter Stegmaier Levain, A., Joly, P.-B., Barbier, M., Cardon, V., Dedieu, F. and Pellissier, F. (2015) Continuous discontinuation – The DDT Ban revisited. International Sustainability Transitions Conference: Sustainability transitions and wider transformative change, historical roots and future pathways. Leyshon, A. (2009) The software slump: Digital music, the democratisation of technology, and the decline of the recording studio sector within the musical economy. Environment and Planning A, 41, 1309–1331. https://doi.org/doi: 10.1068/a40352. Liersch, C.C. and Stegmaier, P. (2022) Keeping the forest above to phase out the coal below: The discursive politics and contested meaning of the Hambach Forest. Energy Research & Social Science, 89, 1–19. Lifshitz-Assaf, H. (2018). Dismantling knowledge boundaries at NASA: The critical role of professional identity in open innovation. Administrative Science Quarterly, 63(4), 746–782. Lindenberg, M. (1989) Making economic adjustment work: The politics of policy implementation. Policy Sciences, 22, 359–394. Lowry, W.R. (2005) Policy reversal and changing politics: State governments and dam removals. State Politics and Policy Quarterly, 5(4), 394–419. Margulles, N. and Raia, A.P. (1967) Scientists, engineers, and technological obsolescence. California Management Review, 10(2), 43–48. Markard, J., Isoaho, K. and Widdel, L. (2023) Discourses around decline: Comparing the debates on coal phase-out in the UK, Germany and Finland. In Koretsky, Z. et al. (eds) Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation. Routledge. Martiskainen, M., et al. (2021) User innovation, niche construction and regime destabilization in heat pump transitions. Environmental Innovation and Societal Transitions, 39, 119–140. Matthews, R.C.O., Feinstein, C.H. and Odling-Smee, J. (1982) British Economic Growth 1856–1973: The Post-War Period in Historical Perspective. Oxford University Press. McCarthy, M.A. (2017) Dismantling Solidarity: Capitalist Politics and American Pensions since the New Deal. Cornell University Press. Meadows, D. H., Meadows, D. L., Randers, J., & Behrens, W. W. (1972) The Limits to Growth. A Report for the Club of Rome’s Project on the Predicament of Mankind. Universe. Melby, L. and Toussaint, P. (2016) “We walk straight past the screens”: The power of the non-user of a hospital information system. In Hyysalo, S., Jensen, T.E. and Oudshoorn, N. (eds) The New Production of Users: Changing Innovation Collectives and Involvement Strategies. Routledge. Mellal, M.A. (2020) Obsolescnece – A review of the literature. Technology in Society, 63, 101347. Middleton, R. (2006) Review article: The political economy of decline. Journal of Contemporary History, 41(3), 573–586. Mohamad, A., et al. (2015) Business discontinuity among small and medium enterprises. Advanced Science Letters, 21(6), 1763–1766. Mokosch, G. (2019) Breaking the spiral: Institutions and Italy’s economic malaise. The International Spectator: Italian Journal of International Affairs, 54(3), 153–155. Mone, M.A., McKinley, W. and Barker, V.L. (1998) Organizational decline and innovation: A contingency framework. The Academy of Management Review, 23(1), 115–132. Myskja, B.K., Nydal, R., Myhr, A.I. (2014) We have never been ELSI researchers – there is no need for a post-ELSI shift. Life Sciences, Society and Policy, 10(9), 1–17. Neville, K.J. (2020) Shadows of divestment: The complications of diverting fossil fuel finance. Global Environmental Politics, 20(2), 3–11. Ober, E., et al. (2017) Planned obsolescence: The government’s choice?PLATE conference, Delft University of Technology, 8–10 November. Conceptual aspects of discontinuation governance 113 Oudelaar, N. (2015) The Influence of Non-Governmental Organisations on European Union Policy Processes – An Explorative Study on the Role of Non-Governmental Organisations during the Policy Process that Went Prior to the Incandescent Light Bulb Phase-Out (Master Thesis, Enschede, University of Twente). Oudshoorn, N. and Pinch, T. (2003) Introduction: How users and non-users matter. In Oudshoorn, N. and Pinch, T. (eds) How Users Matter: The Co-Construction of Users and Technology. MIT Press. Paech, N. (2013) Economic growth and sustainable development. In Angrick, M. et al. (eds) Faktor X, Re-source – Designing the Recycling Society. Springer. Pantzar, M. and Shove, E. (2010) Understanding innovation in practice: A discussion of the production and re-production of NordicWalking. Technology Analysis & Strategic Management, 22(4), 447–461. Patterson, F., Kerrin, M., and Gatto-Roissard, G. (2009) Characteristics & Behaviours of Innovative People in Organizations. City University. Pellissier, F. (2021) Tuer les pestes pour protéger les cultures. Sociohistoire de l’administration des “pesticides” en France (Doctoral dissertation, Université Gustave Eiffel). Pierson, P. (1994) Dismantling the Welfare State?Cambridge University Press. Pike,A.,Coombes,M.,O’Brien, P. and Tomaney, J. (2018) Austerity states, institutional dismantling and the governance of sub-national economic development: The demise of the regional development agencies in England. Territory, Politics, Governance, 6(1), 118–144. Princen, T., et al. (eds) (2015) Ending the Fossil Fuel Era. MIT Press. Prontera, A. (2021) The dismantling of renewable energy policy in Italy. Environmental Politics, 30(7), 1196–1216. Rip, A. (2009) Futures of ELSA. EMBO Reports, 10(7), 666–670. https://doi.org/10.1038/ embor.2009.149 Rip, A. (2012) The context of innovation journeys. Creativity and Innovation Management,21 (2), 158–170. Rip, A. (2018) Futures of Science and Technology in Society. Springer VS. Rip, A. and Kemp, R. (1998) Technological change. In Rayner, S. and Malone, E.L. (eds) Human Choice and Climate Change 2. Battelle Press. Rip, A., Misa, T.J. and Schot, J. (eds) (1995) Managing Technology in Society: The Approach of Constructive Technology Assessment. Pinter. Sabatier, V., et al. (2012) When technological discontinuities and disruptive business models challenge dominant industry logics: Insights from the drugs industry. Technological Forecasting & Social Change, 79, 949–962. Sato, H. (2002) Abolition of leprosy isolation policy in Japan: Policy termination through leadership. Policy Studies Journal, 30(1), 29–46. Sato, H. and Frantz, J.E. (2005) Termination of the leprosy isolation policy in the US and Japan: Science, policy changes, and the garbage can model. BMC International Health and Human Rights, 5(3), 1–16. Sato, H. and Narita, I. (2003) Politics of leprosy segregation in Japan: The emergence, transformation and abolition of the patient segregation policy. Social Science & Medicine, 56, 2529–2539. Schot, J. (1992) Technology in decline: A search for useful concepts. British Journal for the History of Science, 25(1), 5–26. Schot, J. and Geels, F. (2008) Strategic niche management and sustainable innovation journeys: Theory, findings, research agenda, and policy. Technology Analysis and Strategic Management, 20(5), 537–554. Schot, J. and Steinmueller, W.E. (2018a) Three frames for innovation policy: R&D, systems of innovation and transformative change. Research Policy, 47, 1554–1567. https://doi.org/ 10.1016/j.respol.2018.08.011. Schot, J. and Steinmueller, W.E. (2018b) New directions for innovation studies: Missions and transformations. Research Policy, 47, 1583–1584. Shove, E. and Walker, G. (2007) Commentary: CAUTION! Transitions ahead: Politics, practice, and sustainable transition management. Environment and Planning A, 39, 763–770. Shove, E. (2012) The shadowy side of innovation: unmaking and sustainability. Technology Analysis & Strategic Management, 24(4), 363–375. Smith, A. and Kern, F. (2009) The transitions storyline in Dutch environmental policy. Environmental Politics, 18(1), 78–98. Smith, A. and Stirling, A. (2010) The politics of social-ecological resilience and sustainable socio-technical transitions. Ecology and Society, 15(1), 11. Stachowiak, S. (2013) Pathways for change: 10 theories to inform advocacy and policy change efforts. ORS Impact. www.syrialearning.org/system/files/content/resource/ files/main/2013-10-ors-10-theories-to-inform-advocacy-and-policy-change-efforts.pdf Star, S.L. (2010) This is not a boundary-object. Revue d’anthropologie des connaissances, 41(1), 18–35. Stegmaier, P. (2023) Aftercare, or doing less with discontinuation niche governance. In Goulet, F. and Vinck, D. (eds) Doing Without, Doing With Less: The New Horizons of Innovation. Edward Elgar. Stegmaier, P. (2017) The reflexive creation of exiting: The governance of discontinuation. Summer School “Perspectives on Innovation Society”. TU Berlin, 17 June. Stegmaier, P., Kuhlmann, S. and Visser, V. R. (2012a) Governance of the discontinuation of socio-technical systems. Jean Monnet Conference: The Governance of Innovation and Socio-Technical Systems in Europe: New Trends, New Challenges, International Workshop, Copenhagen Business School, Denmark. Stegmaier, P., Visser, V. R. and Kuhlmann, S. (2012b) Governance of the discontinuation of socio-technical systems—An exploratory study of the incandescent light bulb phaseout. 4S/EASST Conference, Panel on the Governance of Innovation and Sociotechnical Systems: Design and Displacements –I, Copenhagen Business School, Denmark. Stegmaier, P. and Kuhlmann, S. (2016) Capturing how obsolescent innovation must give way: The discontinuation governance heuristic and the phase-out of the incandescent light bulbs in the EU). STePS Annual Research Days, University of Twente, Enschede, 14 June. Stegmaier, P., Kuhlmann, S. and Visser, V. R. (2014) The discontinuation of socio-technical systems as governance problem. In Borrás, S. and Edler, J. (eds) Governance of Systems Change. Edward Elgar. Stegmaier, P., Visser, V. R. and Kuhlmann, S. (2021) The incandescent light bulb phaseout. Exploring patterns of framing the governance of discontinuing a socio-technical regime. Energy, Sustainability and Society, 11, 1–22. https://doi.org/https://doi.org/10. 1186/s13705-021-00287-4. Stegmaier, P., Joly, P.-B., Kuhlmann, S. and Stirling, A. (2023) Technologies of Discontinuation –Towards Transformative Innovation Policies. Edward Elgar. Stephenson, J.R., et al. (2021) Energy cultures and national decarbonisation pathways. Renewable and Sustainable Energy Reviews, 137, 110592. Stirling, A. (2014) Transforming power: Social science and the politics of energy choices. Energy Research & Social Science,1,83–95. https://doi.org/https://doi.org/10.1016/j.erss. 2014.02.001. Strauss, A. (1978) Negotiations: Varieties, Contexts, Processes, and Social Order. Jossey-Bass. 114 Peter Stegmaier Conceptual aspects of discontinuation governance 115 Streeck, W. and Thelen, K. (eds.). (2005) Beyond Continuity: Institutional Change in Advanced Political Economies. Oxford University Press. Stuart, R.C. and Panayotopoulos, C.M. (1999) Decline and recovery in transition economies: The impact of initial conditions. Post-Soviet Geography and Economics 14(4), 267–280. Sveiby, K. et al. (2012) Challenging the Innovation Paradigm. Routledge. Tomlinson, J. (2000) The Politics of Decline: Understanding Post-War Britain. Routledge. Turnheim, B. (2023) Destabilisation, decline and phase-out in transitions research. In Koretsky, Z. et al. (eds) Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation. Routledge. Turnheim, B. and Geels, F. (2012) Regime destabilisation as the flipside of energy transitions: Lessons from the history of the British coal industry (1913–1997). Energy Policy, 50, 35–49. Turnhout, E. (2009) The rise and fall of a policy: Policy succession and the attempted termination of ecological corridors policy in the Netherlands. Policy Sciences, 42, 57–72. van Asselt, H. (2014) Governing the transition away from fossil fuels: The role of international institutions. Stockholm Environment Institute Working Paper no. 2014-07. Stockholm Environment Institute – Oxford Centre. van de Graaf, H. and Hoppe, R. (1996) Beleid en Politiek: Een Inleiding tot de Beleidswetenschap en de Beleidskunde. Coutinho. van de Leemput, D. and van Lente, H. (2023) Caring for decline: The case of 16mm film artworks of Tacita Dean. In Koretsky, Z. et al. (eds) Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation. Routledge. van Leuven, A.J. and Hill, E.W. (2021) Legacy regions, not legacy cities: Growth and decline in city-centered regional economies. Journal of Urban Affairs. https://doi.org/10. 1080/07352166.2021.1990775. Villalonga, B. and McGahan, A.M. (2005) The choice among acquisitions, alliances, and divestitures. Strategic Management Journal, 26, 1183–1208. Vogelsangs-Coombs, V. and Keller, L.F. (2013) The government and governance of Ohio: Party politics and the dismantling of public administration. Administration & Society, 47(4), 343–368. Warmington, A. (1974) Obsolescence as an organizational phenomenon. Journal of Management Studies, 11(2), 96–114. Weber, M. (1978). Economy and Society: An Outline of Interpretive Sociology (Vol. 2). University of California Press. Weiner, K. and Will, C. (2016) Users, non-users and “resistance” to pharmaceuticals. In Hyysalo, S., Jensen, T.E. and Oudshoorn, N. (eds) The New Production of Users: Changing Innovation Collectives and Involvement Strategies. Routledge. Weitzel, W. and Jonsson, E. (1989) Decline in organizations: A literature integration and extension. Administrative Science Quarterly, 34(1), 91–109. https://doi.org/www.jstor.org/ stable/2392987. Wyatt, S. (2003) Non-users also matter: The construction of users and non-users of the internet. In Oudshoorn, N. and Pinch, T. (eds) How Users Matter: The Co-Construction of Users and Technology. MIT Press. Yesley, M.S. (2008) What’s ELSI got to do with it? Bioethics and the Human Genome Project. New Genetics and Society, 27(1), 1–6. Zwart, H. and Nelis, A. (2009) What is ELSA Genomics? Profile and challenges of an emerging research practice. EMBO Reports, 10, 540–544. PART II Empirical explorations 5 DISCOURSES AROUND DECLINE Comparing the debates on coal phase-out in the UK, Germany and Finland Jochen Markard, Karoliina Isoaho and Linda Widdel 5.1 Introduction Decline of unsustainable practices is a crucial process in sustainability transitions (Markard and Rosenbloom 2020; Rosenbloom and Rinscheid 2020). Only if problematic practices or technologies decline, and eventually vanish, is there a chance for ecosystems to recover and for socio-technical systems to become more sustainable. To tackle climate change, for example, massive reductions in the use of fossil fuels are required (IPCC 2022). If we want to remain within the 1.5°C target, power generation from coal and natural gas has to decline worldwide at a speed for which there are hardly any historic precedents (Vinichenko et al. 2021). Decline can happen without policy intervention (e.g., DVDs being replaced by video streaming) but it may also be guided by public policy. An example for the latter is the policy driven phase-out and ban of incandescent light bulbs (Stegmaier et al. 2021). Especially when time is running out and negative consequences accumulate as in the case of climate change, it is essential for policymaking to act swiftly and to accelerate processes of decline. A key approach to guide and accelerate decline is phase-out policies: ‘governance interventions aimed at terminating specific technologies, substances, processes, or practices that are considered harmful’ (Rinscheid et al. 2021: 27). Phase-out policies may specify a date by which the practice has to end, a path or steps toward that end, compensations for those negatively affected by the phase-out, and other details (ibid.). Phase-out policies have been implemented for toxic substances such as DDT (Maguire and Hardy 2009), products such as light bulbs (Stegmaier et al. 2021) or gasoline vehicles (Meckling and Nahm 2019) and technologies such as nuclear power (Markard et al. 2020). Phase-out policies are typically very much contested, which is why the underlying politics, i.e., the political processes leading to phase-out, are very important DOI: 10.4324/9781003213642-5 This chapter has been made available under a CC-BY-NC-ND 4.0 license. UK Germany Finland Power generation 335 (2018) 574 (2020) 69 (2020) (TWh) Coal phase-out Nov. 2015 (pledge) Jan. 2019 Apr. 2018 (pledge) decision (commission) Feb. 2019 July 2020 (Parliament) (Parliament) Target date 2025 (now: 2024) 2038 2029 Coal TIS Share of coal at 23% of power 28% of power 14% of power time of pledge generation generation generation Particularities Old, inefficient coal Mostly state-of-the- Most plants from plants, many were art, several new 1980s and 90s, not meeting future plants built recently some newer; emission standards coal for district heating TIS on domestic coal Last major wave of Lignite mining with No domestic coal mining decline in domestic about 20,000 workmining coal mining in the ers; black coal mining 1980s ended in 2018 Competing TIS Renewables, gas, Renewables, gas; Renewables, gas, nuclear nuclear phase-out nuclear TABLE 5.1 Key characteristics of electricity supply systems, coal, phase-out and related TIS 126 Jochen Markard et al. 5.4 Methods and data Our analyses begin in 2000 for all three countries. At this time, concerns over climate change had been clearly formulated but coal phase-out was not an issue yet. The analyses end after the phase-out decisions had been made. For the UK, we collected data until the end of 2017, where we saw that the discourse cooled off after the decision. For Finland and Germany, we therefore set the cut-off date in the month or shortly after the phase-out passed parliament: April 2019 for Finland and July 2020 for Germany. Our results for the UK (Isoaho and Markard 2020) and Germany (Markard et al. 2021) have already been published. Here, we add the comparative perspective as well as unpublished data from Finland. 5.4.1 Methodological background In our study, we analyze the public discourse as it is expressed in newspaper articles. Following earlier studies (e.g. Isoaho and Markard 2020; Leipprand et al. 2016; Rosenbloom et al. 2016), we build on Hajer’s (1995, 2006) argumentative approach to discourse analysis. This approach grants arguments on a specific topic, brought forward by stakeholders, a key role in the political process. Through discourse, shared understandings of social and physical phenomena are created and conflicting views become apparent. A central concept in discourse analysis is the Discourses around decline 127 storyline: ‘a condensed statement summarizing complex narratives, used by people as “short hand” in discussions’ (Hajer 2006: 69). ‘Through storylines, actors select certain aspects of the discourse while excluding others, thereby reducing the complexity of policy issues … For example, the storyline “Coal is bad for the climate” condenses complex arguments on the release of CO 2 through the burning of coal, how CO 2 contributes to climate change and how climate change is bad for society and nature’ (Markard et al. 2021: 317). Discourse analysis has the capacity to reveal conflicts and argumentative reactions, in which some storylines are included and others omitted from the statements of actors (Isoaho and Karhunmaa 2019). The approach also acknowledges that actors’ discursive positions are not constant but subject to change. Therefore, it allows the tracing of changes in the discourse over time. Building on Hajer’s concept of storyline, we take a quantitative approach to analyzing discourse, i.e., we count how often certain storylines were mentioned. 6 We do this to facilitate comparability across the three cases. The downside of this approach is that we provide less information on the (qualitative) content of storylines and how they were presented. Our analysis uses archival data from nationwide newspapers. We focus on newspapers as data source because they are a central medium in which political arguments are exchanged; newspapers, and media more generally, can be viewed as a major environmental policymaking arena alongside more formal venues such as parliamentary debates (Boykoff and Boykoff 2007; Hansen 2010). Moreover, investigating news articles allows us to cover the discourses from both incumbent and niche actors. The media is often interested in highlighting conflicts and struggles to attract attention, and so newspaper articles are likely to contain more diverse actor interests than for example policy documents (Delshad and Raymond 2013). 5.4.2 Data sources and data collection For the UK case study, we downloaded newspaper articles from the LexisNexis academic database. After a scoping phase with test-runs on several newspapers, The Guardian was chosen as the main source as it was the only nationwide quality newspaper available in the database that systematically covered energy and climate issues. 7 For Germany, we chose the daily editions of Süddeutsche Zeitung (SZ) and Die Welt as data sources. Both newspapers report about the German energy market on a regular basis and were accessible through LexisNexis and the Bavarian State Library’s online archive. 8 Both newspapers have different ideological stances: SZ represents center-left and Die Welt conservative, market-liberal values. For the Finnish case study, we chose Helsingin Sanomat, the only nationwide newspaper in Finland, and news articles from YLE, Finland’s nationwide public broadcasting company. 9 Both sources claim to be politically independent and liberally oriented (Teräväinen 2014). To collect relevant articles, we did both a general search for articles on coal used for power generation and a more specific search on coal phase-out. 10 After obtaining the data, we went through all articles and eliminated UK Germany Finland Source(s) The Guardian Süddeutsche Zeitung Helsingin Sanomat (SZ), (HS), Die Welt YLE Time period 2000 – Dec 2017 2000 – July 2020 2000 – April 2019 Number of articles 249 329 (SZ) 77 (HS) 281 (Die Welt) 91 (YLE) Storylines coded 471 814 548 TABLE 5.2 Overview of data sources 128 Jochen Markard et al. duplicates and false positives (e.g., articles about domestic politics, housing, manufacturing etc.). The final data set for the UK consisted of 249 articles (2000–2017), the German data set included 610 articles (2000–2020) and the Finnish 168 articles (2000–2019). See Table 5.2 for an overview. 5.4.3 Data analysis We developed a common strategy for the analysis of storylines. First, two authors 11 examined a subset of articles (every second or third, depending on the sample size) and inductively derived storylines from the sample. Here we followed pre-defined steps: 1) identify key text passages where coal is discussed, 2) code value judgments or arguments related to coal, 3) identify and code actors that make these arguments. After independent analysis, the results were then compared and discussed between the authors and finally consolidated into a list of storylines. Due to this bottom-up approach, storylines vary across countries. However, we also found many similarities (Table 5.3). Next, the coding rules for each storyline were discussed and aligned. Then the entire data set was coded. All coding was performed using the NVivo software package for qualitative analysis. In a final step, similar storylines were grouped together across countries and paraphrased. Table 5.3 lists all storylines sorted by their frequency. Actors were first coded by their name and later assigned to actor groups. The groups were created inductively and updated as the analysis went on. Actor groups were aligned across the three case studies. 5.5 Results We look into four different aspects of the discourses: general dynamic, the most prominent storylines, the most prominent actor groups and which group argues in favor of or against coal phase-out. 5.5.1 Discourse dynamics We find several similarities but also differences in how the discourses unfolded over time in the three countries (Figure 5.3 and Table 5.4). In all countries, we see a clear Storyline (and paraphrase) UK Germany Finland Bad for climate 200 337 171 Coal is bad for the climate. Coal not needed 56 56 161 Coal can be replaced by other energy carriers, it is not needed to secure power supply. Coal is reliable 38 101 43 Coal is needed to keep the lights on. CCS is a solution 75 32 23 CCS technology is a solution to the climate problem, it can capture the emissions. Bad for economy -68 Coal decline and phase-out have a negative effect on jobs and / or regional economies. Structural change needed -52 Structural change is inevitable and there are new jobs in sustainable industries. Coal is cheap 11 46 16 Not so bad -27 Our emissions from coal do not matter at a global scale. Health risk 45 14 - Burning coal pollutes the air and creates health risks. Coal is expensive -38 CCS no solution 29 13 7 CCS is too expensive and risky. Coal is our identity 17 - - We are a coal country, coal is part of who we are. Coal ban problematic -- A ban leaves little time to develop alternatives. No health risk - 5 - Coal is not a health risk. 12 2 6 3 1 4 2 1 6 TABLE 5.3 Overview of storylines and number of codes Discourses around decline 129 peak in media attention around the time when the coal phase-out was announced (or decided) and lively discourse activity in the years before. 12 While there were some sporadic articles on the topic in the early 2000s, more regular discourse activity started around 2007 (UK) or 2008 (Germany, Finland). In the UK we find two waves of attention, with a first major build-up around 2008–2009. These years are characterized by the implementation of the Climate Change Act (2008) and a specificdebatearound carbon capture and storage technology (CCS) as a means to decarbonize new coal power plants (Isoaho and Markard 2020). This debate largely disappeared in subsequent years as CCS turned out not to be viable. In Germany, we find a more or less 130 Jochen Markard et al. FIGURE 5.3 Discourse activity over time per country Share of storylines in favor of phase-out (light grey, positive) and against (dark grey, negative). To calculate the share, we divided the number of storylines in a specific year by the overall number of storylines for the entire period for each country. The columns for the final year in Germany and Finland are shaded because data collection stopped in July and April, respectively. Discourses around decline 131 FIGURE 5.3 (Cont.) steady build up with high discourse activity around the phase-out decision. In 2008, there was also a debate around CCS technology and announcements of pilot plants. In Finland, we see a steady but low discourse activity from 2008–2014, then a first peak in 2015 when there was an intense debate about a new coal power plant in Helsinki, then decreasing attention until a major peak in 2018, when the government proposed to phase-out coal. The timespan from when discourses became more intense until the phase-out decision was about 9 years in the UK (2007–2015), 11 years in Finland (2007– 2018) and more than 12 years in Germany (2008–2020). Comparing storylines in favor (displayed as positive numbers) and against phase-out (negative numbers), we also find commonalities. In all three countries, pro phase-out storylines outnumber contra storylines. As a general development, we see that the early years are characterized by a more equal weight of pro/contra arguments, 13 while in later years, the discourses tilt in favor of pro phase-out arguments. This pattern is most clearly visible in the UK. In Germany, there are two years at the beginning of the debate in which contra arguments had a majority. Contra arguments also remain strong closer to the phase-out decision. The ratio of pro to contra storylines is highest in Finland (2.4), followed by the UK (2.3) and Germany (1.8). The high value in the UK might be an effect of using The Guardian, a left-leaning newspaper. In Germany, the center-left newspaper SZ has a ratio of 2.1 (pro vs contra), while the conservative Die Welt has a value of 1.6. 5.5.2 Which storylines are mobilized? There are many similarities across the three countries with regard to the storylines mentioned (Table 5.3). This is an interesting finding in itself because our bottom- 132 Jochen Markard et al. up approach for identifying storylines (see above) could have led to a very different result. In Figure 5.4, we present six topics in the debate over coal phase-out. For each topic, we list arguments and counter-arguments and how frequently they were mobilized in each country (if at all). We excluded storylines that appeared in one country only. By far the most prominent storyline is that ‘coal is bad for the climate’. It is the cornerstone of the criticism towards coal use in all three countries and it is used in high frequency over the entire time. There is little opposition to this storyline. In Germany and Finland, we see occasional responses arguing that using coal is ‘not so bad’ because there are far bigger emitters elsewhere in the world. We did not find this argument in the UK discourse. A second key topic is about whether coal is needed or not. The most important argument against coal phase-out is about reliability: coal is said to be needed for a stable and secure power supply. This storyline is present in all three countries to a similar extent, although somewhat more frequently in Germany. The counterargument is that ‘coal is not needed’. It is argued, for example, that renewable energy sources can reliably replace it. This storyline was used very often in Finland, but much less so in Germany, where the share of coal for power generation was still rather high at the time of the discussion. Another debate centers around the economic effects of phasing out coal. Some argue that a phase-out is ‘bad for the economy’, e.g., pointing to job losses in German coal mining regions. Others respond that structural changes are needed anyway in these regions and that there are also economic opportunities when developing technological alternatives to coal. These arguments appeared both in Germany and Finland but not in the UK, where jobs in coal mining had already been lost many years before. A fourth topic is about carbon-capture and storage (CCS) technology. Some argue that CCS could be a viable approach to retain the emissions from coal-fired power plants (‘clean coal’), while others were skeptical about the technological and financial viability of CCS and the associated risks. The CCS debate gained prominence in the UK around 2008–2009 and it was also associated with hopes to create a new industry around CCS and to become an international leader (and technology exporter) in this field. CCS was much less of an issue in the other two countries. However, CCS was also discussed in Germany in the same years. Vattenfall Europe built a CCS pilot plant in Eastern Germany 14 and also two other major German utilities, RWE and E.On, argued in favor of CCS technology. In Finland, there was little attention given to CCS at that time. The country saw a CCS debate 10 years later in 2018, when the Finnish utility Fortum announced a pilot project in Norway. Another discussion addressed the costs of coal-fired power generation. Opponents of phase-out argued that coal should be kept because it is cheap, while those in favor of phase-out pointed to the costs that are not accounted for. Despite many similarities, there were also differences between the countries. For example, the argument that there are health risks associated with burning coal was Discourses around decline 133 FIGURE 5.4 Storylines and responses for each country Share of specific storylines in relation to all storylines for the respective country; pro (contra) phase-out: positive (negative) values; no data: storyline was not mentioned. regularly mentioned in the UK but it was hardly mentioned in Germany and not at all in Finland. In addition, we found a pro-coal storyline around ‘coal is part of our national identity’ that was unique for the UK. In Finland, we found a couple of arguments that said a coal ban would be problematic, e.g., as there is no flexibility and it does not leave enough time to develop alternatives. 5.5.3 Which actor groups are present in the discourse? The discourse on coal phase-out is shaped by a broad variety of actors and actor groups. There are six major groups of actors that occur in all three countries, although to varying degrees (Figure 5.5). Overall, environmental NGOs and climate activists present the most dominant actor group. They are by far the most prominent actor group in the UK 15 and also very prominent in Germany. In Finland, however, they were a much less important voice in the debate. A second important group is electric utilities. They were the most active voice in Finland and had an average discourse activity in Germany. In the UK, they were the least prominent group. Scientific experts and think tanks feature prominently and at a similar level in all three countries and so do government actors. Parties played a prominent role in Germany, while they were less important in the two other countries. Industry actors were most prominent in the UK, of some relevance in the German debate and of hardly any importance in Finland. Some groups of actors only appeared in one country. In Germany, we found some federal states (Länder) with strong voices and also labor unions. Both are related to lignite mining, which is concentrated in a few states with jobs at stake. In 134 Jochen Markard et al. FIGURE 5.5 Main groups of actors for each country Share of storylines mentioned by members of a specific actor group in relation to all storylines in the respective country. Finland, cities also had a voice in the debate. Here, most coal power plants are owned by municipal utilities. The plants supply cities with electricity and heat (district heating). In summary, there are fewer similarities in terms of actor groups across countries than in terms of storylines. In other words, similar arguments and counter-argu- ments are made—in part by the same groups of actors, in part by other groups. 5.5.4 Which actor groups mobilize which storylines? Not surprisingly, environmental NGOs and climate activists were the key groups of actors that argued in favor of coal phase-out (Figure 5.6). This pattern holds across all countries, even though NGOs had a much stronger voice in the UK than in Germany and Finland (see above). The main argument of these actors was about climate change. They also argued that coal is not needed. In the UK and Germany, some NGOs were in favor of CCS at the beginning. Science experts and think tanks also spoke mostly in favor of phase-out. Their main arguments were climate change, alternative energy supply options (coal not needed) and, to a lesser extent, that structural change is needed anyway and that there are also economic opportunities in coal phase-out. In addition, there were several voices that saw a merit in CCS technology, especially in the UK and Finland. Both of these groups were quite homogeneous in their positions and did not change their arguments over time (except for the pro CCS arguments). Government actors, in contrast, were less homogeneous. We find, for example, Discourses around decline 135 environment ministers often arguing in favor of coal phase-out, while other government officials argued in favor of CCS (especially in the UK) or raised issues around job losses and potential negative economic impacts of phase-out (primarily Germany and Finland). Also, storylines from government actors shifted over time, from pro-coal to pro-phase-out. Utility companies, coal industry and unions, as well as industry actors, show somewhat contrasting positions to the aforementioned groups. They mobilized more storylines against coal phase-out than in favor and their two most central arguments were security of supply (utilities and coal industry), job losses (coal industry) and costs of supply, i.e. coal being cheap (primarily industry). This pattern is more pronounced in Germany and the UK than in Finland, where coal industry and coal-related labor unions did not play a role in the debate. In Finland, we also found many utilities in favor of phase-out, especially in the year when it was finally decided. A more detailed analysis of the main opposing storylines (Figure 5.7) shows several similarities between the UK and Germany: climate change as the main argument pro phase-out, primarily mobilized by e-NGOs, science actors and parts of government, plus ‘coal not needed’ as an additional but much less prominent argument. In Finland, both climate and ‘not needed’ are of equal importance with all four actor groups mobilizing them (but e-NGOs comparably less). On the ‘procoal’ side we find reliability as the key storyline in all three countries, mostly used by utilities. The UK is an exception here with CCS playing a major role (see above) and the government having quite a prominent voice. In fact, the UK government was the main actor defending a continued use of coal in the early years of the UK discourse. FIGURE 5.6 Main actor groups and whether they were in favor of or against coal phaseout Share of storylines mentioned by members of a specific actor group (ranked from pro to contra phase-out) in relation to all storylines in the respective country. Pro (contra) phase-out: positive (negative) values. 238 Adrian Rinscheid et al. regulatory instruments for phase-outs targeting fossil fuels in specific industries such as construction (Williamson & Finnegan 2021) or fossil fuel based industries and infrastructures in their entirety (Furnaro 2021). Much of this recent work is explicitly prescriptive in calling for command-and-control approaches to phase-out carbon-intensive arrangements domestically, be it coal-fired power generation (Kittel et al. 2020), cement manufacturing (Feng et al. 2018) or fossil fuels in general (Murombo 2021). This prescriptive turn is likely due to a combination of increasing perceptions of climate urgency among scholars (Wilson & Orlove 2021) and frustrations around the fact that other policy instruments like international agreements and economic instruments have largely failed to curtail the extraction and combustion of fossil fuels in line with internationally agreed temperature targets (Piggot et al. 2020). Management and planning approaches (Persson 2006) are also used to drive phase-outs. These rely less on authority and more on the organizational resources of states (see Hood 1983). In the ozone context (11% of instrument mentions), they mostly comprise national-level timetables and targets for ODS phase-outs that deviate from the international ozone regime (e.g., Schmidt 1991). Similarly, the literature sheds light on environmental management tools in the context of healthrelated phase-outs, such as the so-called National Action Plans for reducing the human health and environmental risks of pesticide use, which are mandated by a directive for all EU member states (Schulte-Oehlmann et al. 2011). With respect to climate-related phase-outs, management and planning approaches (23% of instrument mentions) capture a variety of initiatives to phase out carbon-intensive arrangements. Some of these refer to existing policy frameworks such as Hong Kong’s Climate Action Plan (Durmaz et al. 2020), under which fossil fuel based technologies are to be phased out. Others refer to scholarly proposals for phaseouts, such as the idea to progressively discontinue the use of internal combustion engine vehicles in Canada (Bahn et al. 2013). Moreover, the Paris Agreement has given rise to demands that nationally determined contributions (i.e., the nonbinding mechanism allowing each country to set its own mitigation and adaptation objectives) should include precise roadmaps and long-term commitments for the phase-out of fossil fuels (Rauner et al. 2020). The three remaining classes of phase-out instruments—economic instruments, voluntary approaches, and education and capacity building—are discussed much less prominently. However, there is interesting variation concerning their association with the three sustainability challenges under scrutiny (Figure 10.6). Economic instruments are almost absent in the phase-out literature about ozone and health, but account for 25% of mentioned policy instruments in the climate context. About half of these mentions relate to the removal or reform of subsidies for carbon-intensive arrangements, covering a variety of geographic and institutional contexts (Burniaux & Chateau 2014; Gençsü et al. 2020), with much of this work being prescriptive. Besides, emission pricing and emission trading systems are frequently theorized to contribute to climate-relevant phase-outs. For instance, Orthofer et al. (2019) show how an ambitious carbon pricing scheme would lead Phase-out as a policy approach 239 to a coal phase-out in South Africa’s power sector. Meanwhile, using state-level data to examine the impact of the Regional Greenhouse Gas Initiative on coal and natural gas consumption in the electricity sector, Yan (2021) identifies this cap- and-trade program as a direct driver of coal and natural gas phase-outs within regulated US states. In general, the higher prevalence of economic ideas in the climate context appears to be in line with the relatively high level of engagement of scholarly communities from economics and business with climate change (see also Hulme et al. 2018). Educational policy instruments and capacity-building approaches are sometimes used as complementary means to support the implementation of primary phase-out instruments. They play a non-negligible role in the context of ozone (10% of instrument mentions) which reflects scholarly interest in various flanking measures that support the global ODS phase-out. For example, scholarship discusses domestic schemes for technology cooperation and transfer, as well as the Regional Networks of Experts, an institution established by the United Nations Environment Programme (UNEP) to “provide a regular forum for the exchange of knowledge and experience, as well as development of skills among officers of National Ozone Units” (Shende 2015). Education and capacity building also appear occasionally in the context of health-related phase-out discussions, but less so in the climate context. In contrast to the ozone and climate cases, discussions of phase-outs targeting human health challenges sometimes evoke voluntary approaches (8% of instrument mentions). This mainly reflects specific voluntary initiatives and corporate policies to phase out hazardous substances such as phthalate plasticizers (Nardelli et al. 2015) or polybrominated diphenyl ethers (Shaw et al. 2010). By voluntarily committing to a phase-out, a firm can try to position itself as a sustainability champion, based on the calculation to gain a competitive edge (e.g., Paska 2010). Alternatively, voluntary approaches may emerge in an effort to address voids left by missing or weak regulatory instruments (e.g., Tickner et al. 2005). Yet, the role of voluntary approaches has not been examined more systematically so far. To conclude, our results demonstrate that multiple policy instruments are considered as drivers of phase-outs in practice and scientific discourse. Furthermore, patterns can be distilled by examining the specific phase-out instruments associated with different sustainability challenges. Accordingly, discussions of ODS phase-outs are associated mainly with international and global governance instruments, while phase-outs concerning human health issues are more frequently associated with command-and-control approaches. Finally, phase-outs in the context of climate change mitigation tend to be associated with a broader variety of instruments. 10.3.5 Phase-out functions Most scholarly work discusses the phase-out of socio-technical elements primarily as a way to mitigate or ameliorate certain environmental sustainability challenges. Beyond that, a number of publications offer broader reflections on the ways in 240 Adrian Rinscheid et al. which phase-out may shape the evolution of socio-technical systems. Inspired by research on technological innovation systems (Kivimaa & Kern 2016), we refer to these as “functions”; i.e., processes caused by phase-outs that implicate a certain development within socio-technical systems. These are frequently discussed as a rationale for phase-outs. While we do not make the claim to comprehensively and accurately “measure” such functions in academic publications, our inductive coding led to the identification of three distinct sets of logics guiding phase-outs: substituting, signaling and transforming. Of these, substituting is associated with the smallest degree of change in a sociotechnical system. It captures the idea of using a phase-out intervention to remove an undesirable element and replace it with an alternative (Rinscheid et al. 2021). Signaling, by contrast, relates to the notion of using phase-out as a long-term direction-setting device for socio-technical change, simultaneously delegitimizing ongoing investment in the established system and reinforcing the general characteristics of the ultimate end point of a transition (Meckling & Nahm 2019). It emphasizes the role of phase-out as a trigger for innovation in terms of novel technologies, business practices and potentially societal arrangements more broadly. Finally, transforming captures the idea that phase-outs can target a broader range of technologies, infrastructures or even institutions such as entire fossil fuel based industries, often combined with additional policy interventions. In this case, phaseout is used as a lever to incite transformative change across the broader web of interlinked technologies, practices and rules that comprise socio-technical systems (Semieniuk et al. 2021). In the surveyed literature, substituting is evoked regularly across the three challenges studied above (ozone, climate change and human health). It figures most prominently in the context of ozone, where several works put forward the phaseout of ODS as a means to trigger a relatively seamless switch from processes or technologies containing these substances to environmentally more benign replacements (e.g., Agrawal & Shrivastava 2010). The same logic is also evoked in several works relating to health and climate challenges, for instance in the context of replacing coal-fired power plants with natural gas. Signaling and transforming, by contrast, are absent from the literature on ODS phase-out (see Figure 10.7). In some health- and climate-related publications, however, the function of signaling is evoked. Here, it is sometimes discussed as a way to provide sufficient preparation time to industry and stakeholders, aiming to reduce the cost and pain of transitioning from incumbent technologies, materials and business models to more sustainable alternatives (Howarth & Rosenow 2014; Meckling & Nahm 2019). A prototypical contribution to understanding the role of signaling in phase-outs is Rogge and Johnstone’s (2017) paper. This work postulates—and demonstrates empirically—that phase-out policies in the energy sector cause manufacturers’ innovation expenditures for new technologies to rise, outpacing other political interventions by far. Similarly, Bretschger and Zhang (2017) argue that a nuclear phase-out induces innovation and increases investment in renewable energies, an argument reiterated by Carrara (2020). While these Phase-out as a policy approach 241 FIGURE 10.7 Functions and motivations ascribed to phase-outs in the context of three environmental sustainability challenges The horizontal axes indicate the share of publications evoking a function or motivation within the three subsamples. contributions emphasize the key role of policy signals in shaping market actors’ expectations about visions for future arrangements that may add up to systemic change, signaling does not necessarily entail destabilization of incumbent regimes. Rather, through processes like creative accumulation (Bergek et al. 2013), recombination and diversification (Andersen & Gulbrandsen 2020), incumbents may be able to internalize a phase-out and remain among the shapers of socio-technical systems. Transforming presumes far-reaching implications of phase-out in that such interventions can unsettle incumbent systems, trigger or accelerate broader change processes, and thereby fundamentally transform entire socio-technical systems (Rinscheid et al. 2021). An example associated with health challenges involves phase-out interventions targeting the use of toxic materials, thereby advancing the transformation towards a circular economy in the building sector (Wiprächtiger et al. 2020). Similarly, in the climate context, phase-out interventions targeting coal mining or power generation are frequently associated with fundamental transformations towards a sustainable society (e.g., Vögele et al. 2018). Focusing on the phase-out of lignite in Greece, Nikas et al. (2020) consider phase-out as a means to overcome multiple lock-ins and enact multidimensional changes in the Greek economy. Moreover, a number of works emphasize how the removaloffossilfuelsubsidies mayhave far-reaching implications for energy supply systems and foster sustainable development by rupturing the tendency to follow historically established carbon-intensive development paths (Mutanga et al. 2018; Xu 2021). 242 Adrian Rinscheid et al. 10.3.6 Motivations for phase-outs Researchers also observe a variety of motivations for phase-outs (see Figure 10.7). These broadly relate to additional societal drivers of phase-out interventions, alongside the goal of enhancing environmental sustainability. As our unit of analysis is the scientific publication, these motivations sometimes also reflect strategies of legitimization put forward in scholarly reflections about phase-out. Across the three challenges investigated above, the most frequently evoked motivation refers to economic benefits. Yet, the relevance of economic benefits varies strongly. While almost absent from the publications focusing on ozone depletion, they are more relevant in the human health context and even more so in the context of climate change. With respect to human health challenges, phase-out is often discussed as a way to avoid economic costs incurred for society, for instance by nuclear accidents or the accumulation of toxic substances in the biosphere (e.g., De Cian et al. 2014). Regarding climate change, economic benefits ascribed to phase-outs include multifaceted considerations. These range from economic savings (e.g., from phasing out fossil fuel subsidies) to employment creation (e.g., through a transition to decentralized renewable energy), to bolstering energy security, and finally to mitigating financial risks (e.g. stranded asset risks associated with coal-fired power plants). Motivations relating to ethics and equity, which comprise concerns for future generations or social equity more generally, are much less pronounced in the literature. Finally, public demand is discussed as a societal driver for phase-out interventions across all three challenges. While it plays a marginal role in the context of climate change related phase-out discussions, it figures more prominently in particular in relation to human health challenges (e.g., Viklund 2004). 10.4 Discussion and conclusion The purpose of this chapter was to trace how scholarly work has mobilized phaseout, understood as a policy approach to terminate socio-technical elements with harmful or undesirable properties for environmental sustainability outcomes. Based on our systematic review, we derive five main insights. First, after a slow accumulation of early works from the 1970s to the 1990s and a period of markedly higher academic interest between 2006 and 2015, scholarly work on phase-out has expanded rapidly since 2016. Second, different academic communities from the social and natural sciences, engineering, medicine and others contribute to scholarly debates about phase-out. The shared interest in phase-out as a policy approach to address sustainability challenges and its prominence in current policy debates suggest that phase-out serves as a boundary object that mediates between diverse constituencies (Turnheim et al. 2020). It thus connects science with policymakers as much as it does differing strands of science. Third, phase-out has been employed and conceived to tackle a varying set of sustainability challenges over time. Initially associated strongly with the challenge of Phase-out as a policy approach 243 atmospheric ozone depletion, phase-out has also been considered as a tool to tackle human health and other environmental challenges. More recently, phase-outs are increasingly discussed in the context of climate change mitigation. Fourth, across the scholarly publications reviewed, a broad range of policy instruments are described to pursue phase-outs. While the ozone crisis is predominantly addressed with an internationally coordinated approach, human health challenges are often associated with domestic phase-outs using command-and-control instruments. Climate-related phase-outs rely on a range of instruments including command-and-control, international agreements, management and planning, and economic instruments. Fifth, some of the more conceptually oriented literature associates phase-out with specific functions that entail implications for broader socio-technical systems. These range from phase-out as a means for substituting particular elements to phase-out as a signal for a potentially broader societal shift, to phase-out as a vehicle for enacting transformative change. Finally, methodological choices and implications for future research deserve highlighting. A limitation of this investigation into phase-out was to rely on academic publications as units of analysis. Therefore, our findings may be as much a product of academic trends and analytical inclinations as they are a reflection of how phase-out is employed as a policy approach for confronting sustainability challenges. We thus caution against overinterpreting our scientometric analyses. While they reflect trends in the literature, they do not necessarily indicate the main targets and challenges addressed by phase-out policies at a given point in time. Moreover, our analysis does not distinguish between discussions of phase-out as a theoretical concept, an empirical case, a socio-technical process, a unit of a computational model or a normative goal. Another methodological limitation is that our systematic coding relied on the titles, abstracts and keywords of publications, but not the full texts. Future research needs to deepen our knowledge on phase-out, both conceptually and empirically. Our review of 868 publications suggests that theory-building on phase-out has not progressed very far to date. While phase-out has become a major tool in confronting multiple sustainability crises, a widely known and accepted theoretical model of phase-out has not yet been developed. One step in this direction may be seen in our investigation of the functions ascribed to phase-out. Yet, our inductive analysis is not entirely unproblematic, since the functions identified in the literature partly overlap, may not capture the full array of logics guiding phase-outs, and are theoretically underspecified. Still, they may provide a useful starting point for more theoretically informed work on phase-out. Empirically, we urge for systematic comparative analyses of the conditions that allow phase-out interventions to be considered and enacted as a tool for confronting sustainability challenges, the challenges that accompany its implementation in various contexts, and potential coping strategies. Important questions to be addressed include: Which policy instruments or mixes are most effective in phasing out unsustainable socio-technical elements? How can phase-out interventions be designed to ensure acceptance by affected stakeholders and regions without compromising their intended 244 Adrian Rinscheid et al. goals? How do phase-out interventions interact with other policy approaches steering innovation and decline in socio-technical systems? And to what extent can insights from “successful” phase-outs be transferred to other settings (i.e., other places, challenges, political contexts, etc.)? In tackling these questions, we hope that different academic communities keep contributing to this field of research, with the aim of generating knowledge that both advances the state of the art and informs decisionmaking for sustainable development. Note 1 We thank Peter Stegmaier, Bruno Turnheim and Joshua Woodyatt for excellent comments on an earlier draft of this chapter. Adrian Rinscheid acknowledges funding from the Basic Research Fund (GFF) at the University of St. Gallen. Daniel Rosenbloom acknowledges the financial support of the Social Sciences and Humanities Research Council of Canada. References Agrawal, A.B. and Shrivastava, V. (2010) Retrofitting of vapour compression refrigeration trainer by an eco-friendly refrigerant. Indian Journal of Science and Technology, 3(4), 455– 458. Andersen, A.D. and Gulbrandsen, M. (2020) The innovation and industry dynamics of technology phase-out in sustainability transitions: Insights from diversifying petroleum technology suppliers in Norway. Energy Research and Social Science, 64, 101447. Andersen, S.O. (2015) Lessons from the stratospheric ozone layer protection for climate. Journal of Environmental Studies and Sciences, 5, 143–162. https://doi.org/10.1007/ s13412-014-0213-9 Andersen, S.O., Halberstadt, M.L. and Borgford-Parnell, N. (2013) Stratospheric ozone, global warming, and the principle of unintended consequences: An ongoing science and policy success story. Journal of the Air and Waste Management Association, 63(6), 607–647. Ashford, N.A. and Hall, R.P. (2019) Technology, Globalization, and Sustainable Development: Transforming the Industrial State. Routledge. Bahn, O., Marcy, M., Vaillancourt, K. and Waaub, J.P. (2013) Electrification of the Canadian road transportation sector: A 2050 outlook with TIMES-Canada. Energy Policy, 62, 593–606. Bergek, A., Berggren, C., Magnusson, T. and Hobday, M. (2013) Technological discontinuities and the challenge for incumbent firms: Destruction, disruption or creative accumulation? Research Policy, 42(6–7), 1210–1224. Blondeel, M., van de Graaf, T. and Haesebrouck, T. (2020) Moving beyond coal: Exploring and explaining the Powering Past Coal Alliance. Energy Research and Social Science, 59, 101304. Bornmann, L., Haunschild, R. and Mutz, R. (2021) Growth rates of modern science: A latent piecewise growth curve approach to model publication numbers from established and new literature databases. Humanities and Social Sciences Communications, 8(1), 1–15. Brauers, H., Oei, P.-Y. and Walk, P. (2020) Comparing coal phase-out pathways: The United Kingdom’s and Germany’s diverging transitions. Environmental Innovation and Societal Transitions, 37, 238–253. Bretschger, L. and Zhang, L. (2017) Nuclear phase-out under stringent climate policies: A dynamic macroeconomic analysis. Energy Journal, 38(1), 167–194. Phase-out as a policy approach 245 Burniaux, J.M. and Chateau, J. (2014) Greenhouse gases mitigation potential and economic efficiency of phasing-out fossil fuel subsidies. International Economics, 140, 71–88. Carrara, S. (2020) Reactor ageing and phase-out policies: Global and regional prospects for nuclear power generation. Energy Policy, 147, 111834. De Cian, E., Carrara, S. and Tavoni, M. (2014) Innovation benefits from nuclear phase-out: Can they compensate the costs? Climatic change, 123(3), 637–650. Dewling, R.T. and Anderson, P.W. (1976) New York Bight I: Ocean dumping policies. Oceanus, 19(4), 2–10. Dunn, J.R. (1990) Environmental concerns will widen auto use of specialty elastomers. Elastomerics, 122(2), 12–12. Durmaz, T., Pommeret, A. and Tastan, H. (2020) Estimation of residential electricity demand in Hong Kong under electricity charge subsidies. Energy Economics, 88, 104742. Feng, X.-Z., Lugovoy, O. and Qin, H. (2018) Co-controlling CO2 and NOx emission in China’s cement industry: An optimal development pathway study. Advances in Climate Change Research, 9(1), 34–42. Fraser, P.J. and Bouma, W.J. (1990) New ozone protocol and Australia. Search, 21(8), 261– 264. Furnaro, A. (2021) The role of moral devaluation in phasing out fossil fuels: Limits for a socioecological fix. Antipode, 53(5), 1442–1462. Geels, F.W., Sovacool, B.K., Schwanen, T. and Sorrell, S. (2017) The socio-technical dynamics of low-carbon transitions. Joule, 1(3), 463–479. Gençsü, I., Whitley, S., Trilling, M., et al. (2020) Phasing out public financial flows to fossil fuel production in Europe. Climate Policy, 20(8), 1010–1023. Gonzalez, M., Taddonio, K.N. and Sherman, N.J. (2015) The Montreal Protocol: How today’s successes offer a pathway to the future. Journal of Environmental Studies and Sciences, 5(2), 122–129. Grant, M.J. and Booth, A. (2009) A typology of reviews: An analysis of 14 review types and associated methodologies. Health Information and Libraries Journal, 26(2), 91–108. Graziosi, F., Arduini, J., Furlani, F., et al. (2017) European emissions of the powerful greenhouse gases hydrofluorocarbons inferred from atmospheric measurements and their comparison with annual national reports to UNFCCC. Atmospheric Environment, 158, 85–97. Haddaway, N.R., Woodcock, P., Macura, B. and Collins, A. (2015) Making literature reviews more reliable through application of lessons from systematic reviews. Conservation Biology, 29(6), 1596–1605. Hood, C.C. (1983) The Tools of Government. MacMillan. Howarth, N.A.A. and Rosenow, J. (2014) Banning the bulb: Institutional evolution and the phased ban of incandescent lighting in Germany. Energy Policy, 67, 737–746. Hulme, M., Obermeister, N., Randalls, S. and Borie, M. (2018) Framing the challenge of climate change in nature and science editorials. Nature Climate Change, 8, 515–522. Johnstone, P. and Newell, P. (2018) Sustainability transitions and the state. Environmental Innovation and Societal Transitions, 27, 72–82. Jordan, A., Wurzel, R.K.W. and Zito, A.R. (2003) “New” instruments of environmental governance: Patterns and pathways of change. Environmental Politics, 12(1), 1–24. Kittel, M., Goeke, L., Kemfert, C., et al. (2020) Scenarios for coal-exit in Germany: A model-based analysis and implications in the European context. Energies, 13(8), 2041. Kivimaa, P. and Kern, F. (2016) Creative destruction or mere niche support? Innovation policy mixes for sustainability transitions. Research Policy, 45(1), 205–217. Koretsky, Z. (2023) Dynamics of technological decline as socio-material unravelling. In Koretsky, Z. et al. (eds) Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation. Routledge. 246 Adrian Rinscheid et al. Krause, F., Bach, W. and Koomey, J. (1990) Energy Policy in the Greenhouse. Earthscan. Li, Y., Chang, Q., Duan, H., et al. (2019) Occurrence, levels and profiles of brominated flame retardants in daily-use consumer products on the Chinese market. Environmental Science: Processes & Impacts, 21(3), 446–455. Lovei, M. (1998) Phasing out Lead from Gasoline. World Bank Technical Paper. https://elaw. org/es/system/files/worldbanklead.pdf Markowitz, G. and Rosner, D. (2002) Deceit and Denial: The Deadly Politics of Industrial Pollution. University of California Press. McCulloch, A. (1999) CFC and Halon replacements in the environment. Journal of Fluorine Chemistry, 100(1), 163–173. Meckling, J. and Nahm, J. (2019) The politics of technology bans: Industrial policy competition and green goals for the auto industry. Energy Policy,126, 470–479. Murombo, T. (2021) Regulatory imperatives for renewable energy: South African perspectives. JournalofAfrican Law,66(1),97–122. Mutanga, S.S., Quitzow, R. and Steckel, J.C. (2018) Tackling energy, climate and development challenges in Africa. Economics, 12(1). doi:10.5018/economics-ejournal.ja.2018-61. Nardelli, T.C., Erythropel, H.C. and Robaire, B. (2015) Toxicogenomic screening of replacements for Di(2-Ethylhexyl) Phthalate (DEHP) using the immortalized TM4 Sertoli cell line. PLOS ONE, 10(10), e0138421. Nikas, A., Neofytou, H., Karamaneas, A., et al. (2020) Sustainable and socially just transition to a post-lignite era in Greece: A multi-level perspective. Energy Sources, Part B: Economics, Planning, and Policy, 15(10–12), 513–544. Normann, H.E. (2019) Conditions for the deliberate destabilisation of established industries: Lessons from U.S. tobacco control policy and the closure of Dutch coal mines. Environmental Innovation and Societal Transitions, 33, 102–114. OECD (2001) Sustainable Development: Critical Issues. OECD. Orthofer, C.L., Huppmann, D. and Krey, V. (2019) South Africa after Paris: Fracking its way to the NDCs? Frontiers in Energy Research,7,1–15. Paska, D. (2010) Facilitating substance phase-out through material information systems and improving environmental impacts in the recycling stage of a product. Natural Resources Forum, 34(3), 200–210. Peñasco, C., Anadón, L.D. and Verdolini, E. (2021) Systematic review of the outcomes and trade-offs of ten types of decarbonization policy instruments. Nature Climate Change, 11, 257–265. Persson, Å. (2006) Characterizing the policy instrument mixes for municipal waste in Sweden and England. European Environment, 16(4), 213–231. Piggot, G., Verkuijl, C., van Asselt, H. and Lazarus, M. (2020) Curbing fossil fuel supply to achieve climate goals. Climate Policy, 20(8), 881–887. Rauner, S., Bauer, N., Dirnaichner, A., et al. (2020) Coal-exit health and environmental damage reductions outweigh economic impacts. Nature Climate Change, 10(4), 308–312. Rinscheid, A., Rosenbloom, D., Markard, J. and Turnheim, B. (2021) From terminating to transforming: The role of phase-out in sustainability transitions. Environmental Innovation and Societal Transitions, 41, 27–31. Rogge, K.S. and Johnstone, P. (2017) Exploring the role of phase-out policies for lowcarbon energy transitions: The case of the German Energiewende. Energy Research and Social Science, 33, 128–137. Rosenbloom, D. (2018) Framing low-carbon pathways: A discursive analysis of contending storylines surrounding the phase-out of coal-fired power in Ontario. Environmental Innovation and Societal Transitions, 27, 129–145. Phase-out as a policy approach 247 Rosenbloom, D. and Rinscheid, A. (2020) Deliberate decline: An emerging frontier for the study and practice of decarbonization. Wiley Interdisciplinary Reviews: Climate Change, 11(6), e669. Rotmans, J., Kemp, R. and van Asselt, M. (2001) More evolution than revolution: Transition management in public policy. Foresight,3(1),15–31. Schmidt, K. (1991) How industrial countries are responding to global climate change. International Environmental Affairs, 3(4), 292–315. Schneider, L. (2021) When toxic chemicals refuse to die: An examination of the prolonged mercury pesticide use in Australia. Elementa – Science of the Anthropocene, 9(1), 1–18. Schulte-Oehlmann, U., Oehlmann, J. and Keil, F. (2011) Before the curtain falls: Endocrineactive pesticides – A German contamination legacy. Reviews of Environmental Contamination and Toxicology,213, 137–159. Semieniuk, G., Campiglio, E., Mercure, J.-F., Volz, U. and Edwards, N.R. (2021) Lowcarbon transition risks for finance. Wiley Interdisciplinary Reviews: Climate Change, 12(1), e678. https://doi.org/10.1002/wcc.678 Shaw, S.D., Blum, A., Weber, R., et al. (2010) Halogenated flame retardants: Do the fire safety benefits justify the risks? Reviews on Environmental Health, 25(4), 261–305. Shende, R. (2015) Networking to save the world: UNEP’s regional networks – conflict resolution in action. Journal of Environmental Studies and Sciences, 5(2), 138–142. Stegmaier, P. (2023) Conceptual aspects of discontinuation governance: An exploration. In Koretsky, Z. et al. (eds) Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation. Routledge. Stryker, R.G. (1991) The regulatory phase-out of CFCs and halons. Energy Engineering,88 (3), 22–27. Thornburg, E. (2016) Public as private and private as public: MTBE litigation in the United States. In Hensler, D.R., Hodges, C. and Tzankova, I. (eds) Class Actions in Context. Edward Elgar. Tickner, J., Geiser, K. and Coffin, M. (2005) The U.S. experience in promoting sustainable chemistry. Environmental Science and Pollution Research, 12(2), 115–123. Trencher, G., Rinscheid, A., Rosenbloom, D. and Truong, N. (in review) The rise of phase-out as a critical decarbonisation approach: A systematic review. Environmental Research Letters. Turnheim, B. (2023) Destabilisation, decline and phase-out in transitions research. In Koretsky, Z. et al. (eds) Technologies in Decline: Socio-Technical Approaches to Discontinuation and Destabilisation. Routledge. Turnheim, B., Asquith, M. and Geels, F.W. (2020) Making sustainability transitions research policy-relevant: Challenges at the science–policy interface. Environmental Innovation and Societal Transitions, 34, 116–120. UNFCCC (2021) Glasgow Climate Pact. https://unfccc.int/sites/default/files/resource/cop26_ auv_2f_cover_decision.pdf. van Soest, H.L., de Boer, H.S., Roelfsema, M., et al. (2017) Early action on Paris Agreement allows for more time to change energy systems. Climatic Change, 144(2), 165–179. Velders, G.J.M., Fahey, D.W., Daniel, J.S., et al. (2015) Future atmospheric abundances and climate forcings from scenarios of global and regional hydrofluorocarbon (HFC) emissions. Atmospheric Environment, 123, 200–209. Viklund, M. (2004) Energy policy options – from the perspective of public attitudes and risk perceptions. Energy Policy, 32(10), 1159–1171. Vögele, S., Kunz, P., Rübbelke, D. and Stahlke, T. (2018) Transformation pathways of phasing out coal-fired power plants in Germany. Energy, Sustainability and Society, 8(1), 25. 254 Peter Newman The worst part of this story is that low-income nations continued to use leaded petrol, and in their crowded cities the impacts were terrible. It took another 20 years after the high-income nations’ phase-out for leaded petrol to be phased out globally, with Algeria in 2021 becoming the last country. Lead particles remain along main roads, especially in the slow phase-out cities. As Ritchie said in February 2022: Around one-in-three children globally suffer from lead poisoning… Lead poisoning is estimated to account for about 1% of the global disease burden… in some countries, the costs of lead exposure are equivalent to as much as 6% of GDP. (Ritchie 2022) 11.5 The campaign for phasing out leaded petrol in high- and low-income nations The 1980s were characterised by a growing campaign by newspapers such as The Guardian and The Observer through civil society groups such as The Conservation Society. Much research by medical academics was gathering in the US. The growing weight of the scientific evidence and the horror of leaded petrol’s impact on unsuspecting children began to emerge. The articles and writings from academics outlined the long history of lead from the 1920s, when Thomas Midgley showed Cadillac how adding lead to petrol was a cheap and easy way to prevent engine knocking and “had no side effects” (Needleman 1999). Although leaded petrol use had become widespread, its global acceptance was now coming into conflict with serious health issues. Scientists and civil society began forming associations to fight leaded petrol in the 1980s (Newman and Kenworthy 1985). Elizabeth O’Brien in Sydney established The LEAD (Lead Education and Abatement Design) Group, along with three other families who were dealing with lead poisoning among their relatives. In the LEAD Group Newsletter, the role of academics in providing perspectives now was given an activist outlet, for example, Pollution and the Poor on leaded petrol and social justice (Newman 1993) and Green Accounting and Car-Free Planning (Newman 1998) were distributed widely in the LEAD Group Newsletter. The LEAD Group’s leaded petrol phase-out campaign was part of a wider car use reduction campaign that is now focused on reducing greenhouse gas emissions but from the 1980s had a broad approach to multiple impacts from the over-use of automobiles (Newman and Kenworthy 1992). Although the first steps to transition away from leaded petrol started in the US in 1972 with the governments’ requirement that unleaded petrol be made available, the first country to complete the phase-out of leaded petrol was Japan in 1986 (Lovei 1998). As Japan had also required catalytic converters on cars manufactured for in-country sale, it was not unexpected that they would also put them in their exported vehicles thus ensuring that other countries had to follow by [Document text truncated for crawler view.]