Depuis la révolution industrielle, l'Homme développe des systèmes industriels pour satisfaire ses exigences de production. Mais exploiter ces installations n'est pas sans risques pour les utilisateurs. De ce fait, l'analyse des risques s'est largement développée durant ces dernières décennies. En effet, si dans les années 70, les études se focalisaient sur les défaillances technologiques, des accidents ont souligné l'importance des facteurs humains et organisationnels dans leur occurrence. Si bien que dans les années 80, des méthodes consacrées à l'identification de ces facteurs ont émergées. Ces études, impliquant différentes disciplines, étaient jusqu'alors conçues et conduites séparément les unes des autres. Cet état de fait amène à une sectorisation des analyses et ne permet pas d'avoir une vision globale de la situation étudiée. Mais, depuis peu, des méthodologies proposent d'intégrer (partiellement) ces dimensions dans la démarche d'analyse. Le manque d'intégration constitue aujourd'hui une problématique, scientifique et industrielle, pour les exploitants de systèmes critiques. Ainsi, notre contribution porte sur le développement d'une méthodologie permettant l'analyse de risques de systèmes socio-techniques en exploitation. Ce type d'analyse vise à probabiliser le risque à des fins d'aide à la décision. En ce sens, nous proposons une démarche de formalisation, d'intégration, de représentation et de modélisation des différentes connaissances du système. Le modèle présenté permet d'identifier l'ensemble des causes menant à l'occurrence d'un événement critique, en considérant les données techniques du système et les données liées aux opérateurs et à l'organisation. / Since the industrial revolution, human being develops industrial systems to meet his production needs. But the operation of such facilities involves risks for the users. As a result the risk analysis has expanded during these last decades. Indeed, if in the Seventies, the studies were focused on the technological failures, several major accidents have underlined the importance of human and organisational factors in their occurrence, and have changed this initial way of thinking. So that in the Eighties, different methods allowing an identification of these factors have emerged. These studies, implying different fields of expertise, were so far independently built and applied. This fact leads to sector-based analyses and prevents from having an overall view of the studied situation. But, recently, some methodologies propose to (partially) integrate these different methods to study risks in a global approach. This lack of integration constitutes nowadays a scientific and industrial issue for the owners of critical systems. Thus, our contribution concerns the development of a methodology enabling the risk analyses of socio-technical systems in operation. This kind of analysis aims to probabilistically estimate risks for helping the decision-making. In that way, we propose an approach that enables to formalise, integrate, characterise and represent the different knowledge of the system. Our model allows an identification of the whole of the causes that lead to the occurrence of a critical event, by considering the technical data of the system and the data related to human operators and organisational features.

Depuis la révolution industrielle, l'Homme développe des systèmes industriels pour satisfaire ses exigences de production. Mais exploiter ces installations n'est pas sans risques pour les utilisateurs. De ce fait, l'analyse des risques s'est largement développée durant ces dernières décennies. En effet, si dans les années 70, les études se focalisaient sur les défaillances technologiques, des accidents ont souligné l'importance des facteurs humains et organisationnels dans leur occurrence. Si bien que dans les années 80, des méthodes consacrées à l'identification de ces facteurs ont émergées. Ces études, impliquant différentes disciplines, étaient jusqu'alors conçues et conduites séparément les unes des autres. Cet état de fait amène à une sectorisation des analyses et ne permet pas d'avoir une vision globale de la situation étudiée. Mais, depuis peu, des méthodologies proposent d'intégrer (partiellement) ces dimensions dans la démarche d'analyse. Le manque d'intégration constitue aujourd'hui une problématique, scientifique et industrielle, pour les exploitants de systèmes critiques. Ainsi, notre contribution porte sur le développement d'une méthodologie permettant l'analyse de risques de systèmes socio-techniques en exploitation. Ce type d'analyse vise à probabiliser le risque à des fins d'aide à la décision. En ce sens, nous proposons une démarche de formalisation, d'intégration, de représentation et de modélisation des différentes connaissances du système. Le modèle présenté permet d'identifier l'ensemble des causes menant à l'occurrence d'un événement critique, en considérant les données techniques du système et les données liées aux opérateurs et à l'organisation.

1. Contextual outline of the PhD Research Climate change is today often seen as one of the most challenging issue that our civilisation will have to face during the 21st century. This is especially so now that the most recent scientific data have led to the conclusion that the globally averaged net effect of human activities since 1750 has been one of warming (IPCC 2007, p. 5) and that continued greenhouse gas emissions at or above current rates would cause further warming (IPCC, 2007 p. 13). This unequivocal link between climate change and anthropogenic activities requires an urgent, world-wide shift towards a low carbon economy (STERN 2006 p. iv) and coordinated policies and measures to manage this transition. The climate issue is undoubtedly a typical policy question and as such, is considered amenable to economic scrutiny. Indeed, in today’s world economics is inevitable when it comes to arbitrages in the field of policy making. From the very beginning of international talks on climate change, up until the most recent discussions on a post-Kyoto international framework, economic arguments have turned out to be crucial elements of the analysis that shapes policy responses to the climate threat. This can be illustrated by the prominent role that economics has played in the different analyses produced by the Intergovernmental Panel on Climate Change (IPCC) to assess the impact of climate change on society. The starting point and the core idea of this PhD research is the long-held observation that the threat of climate change calls for a change of climate in economics. Borrowing from the jargon used in climate policy, adaptation measures could also usefully target the academic discipline of economics. Given that inherent characteristics of the climate problem (e.g. complexity, irreversibility, deep uncertainty, etc.) challenge core economic assumptions, mainstream economic theory does not appear as appropriately equipped to deal with this crucial issue. This makes that new assumptions and analyses are needed in economics in order to comprehend and respond to the problem of climate change. In parallel (and without environmental considerations being specifically the driving force to it), the mainstream model in economics has also long been (and still is) strongly criticised and disputed by numerous scholars - both from within and outside the field of economics. For the sake of functionality, these criticisms - whether they relate to theoretical inconsistencies or are empirically-based - can be subsumed as all challenging part of the Cartesian/Newtonian legacy of economics. This legacy can be shown to have led to a model imprinted with what could be called “mechanistic reductionism”. The mechanistic side refers to the Homo oeconomicus construct while reductionism refers to the quest for micro-foundations materialised with the representative agent hypothesis. These two hypotheses constitute, together with the conjecture of perfect markets, the building blocks of the framework of general equilibrium economics. Even though it is functional for the purpose of this work to present them separately, the flaws of economics in dealing with the specificities of the climate issue are not considered independent from the fundamental objections made to the theoretical framework of mainstream economics. The former only make the latter seem more pregnant while the current failure of traditional climate policies informed by mainstream economics render the need for complementary approaches more urgent. 2. Overview of the approach and its main insights for climate policy Starting from this observation, the main objective of this PhD is thus to assess the implications for climate policy that arise from adopting an alternative analytical economic framework. The stance is that the coupling of insights from the framework of evolutionary economics with the perspective of ecological economics provides a promising way forward both theoretically as well as on a more applied basis with respect to a better comprehension of the socioeconomic aspects related to the climate problem. As claimed in van den Bergh (2007, p. 521), ecological economics and evolutionary economics “share many characteristics and can be combined in a fruitful way" - which renders the coupling approach both legitimate and promising. The choice of an evolutionary line of thought initially stems from its core characteristic: given its focus on innovation and system change it provides a useful approach to start with for assessing and managing the needed transition towards a low carbon economy. Besides, its shift of focus towards a better understanding of economic dynamics together with its departure from the perfect rationality hypothesis renders evolutionary economics a suitable theoretical complement for designing environmental policies. The notions of path-dependence and lock-in can be seen as the core elements from this PhD research. They arise from adopting a framework which is founded on a different view of individual rationality and that allows for richer and more complex causalities to be accounted for. In a quest for surmounting the above-mentioned problem of reductionism, our framework builds on the idea of ‘multi-level selection’. This means that our analytical framework should be able to accommodate not only for upward but also for downward causation, without giving analytical priority to any level over the other. One crucial implication of such a framework is that the notion of circularity becomes the core dynamic, highlighting the importance of historicity, feedbacks and emergent properties. More precisely, the added value of the perspective adopted in this PhD research is that it highlights the role played by inertia and path-dependence. Obviously, it is essential to have a good understanding of the underlying causes of that inertia prior to devising on how to enforce a change. Providing a clear picture of the socio-economic processes at play in shaping socio-technical systems is thus a necessary first step in order to usefully complement policy-making in the field of energy and climate change. In providing an analytical basis for this important diagnosis to be performed, the use of the evolutionary framework sheds a new light on the transition towards low-carbon socio-technical systems. The objective is to suggest strategies that could prove efficient in triggering the needed transition such as it has been the case in past “lock-in” stories. Most notably, the evolutionary framework allows us to depict the presence of two sources of inertia (i.e at the levels of individuals through “habits” and at the level of socio-technical systems) that mutually reinforce each other in a path-dependent manner. Within the broad perspective on path dependence and lock-in, this PhD research has first sketched the implications for climate policy of applying the concept of ‘technological lock-in’ in a systemic perspective. We then investigated in more details the notion of habits. This is important as the ‘behavioural’ part of the lock-in process, although explicitly acknowledged in the pioneer work of Paul David (David, 1985, p. 336), has been neglected in most of subsequent analyses. Throughout this study, the notion of habits has been studied at both the theoretical and applied level of analysis as well as from an empirical perspective. As shown in the first chapters of the PhD, the advantage of our approach is that it can incorporate theories that so far have been presented opposite, partial and incomplete perspectives. For instance, it is shown that our evolutionary approach not only is able to provide explanation to some of the puzzling questions in economics (e.g. the problem of strong reciprocity displayed by individual in anonymous one-shot situations) but also is very helpful in bringing a complementary explanation with respect to the famous debate on the ‘no-regret’ emission reduction potential which agitates the experts of climate policy. An emission reduction potential is said to be "no regret" when the costs of implementing a measure are more than offset by the benefits it generates such as, for instance, reduced energy bills. In explaining why individuals do not spontaneously implement those highly profitable energy-efficient investments , it appears that most prior analyses have neglected the importance of non-economic obstacle. They are often referred to as “barriers” and partly relate to the ‘bounded rationality’ of economic agent. As developed in the different chapters of this PhD research, the framework of evolutionary economics is very useful in that it is able to provide a two-fold account (i.e. relying on both individual and socio-technical sources of inertia) of this limited rationality that prevent individuals to act as purely optimising agents. Bearing this context in mind, the concept of habits, as defined and developed in this study, is essential in analysing the determinants of energy consumption. Indeed, this concept sheds an insightful light on the puzzling question of why energy consumption keeps rising even though there is an evident increase of awareness and concern about energy-related environmental issues such as climate change. Indeed, if we subscribe to the idea that energy-consuming behaviours are often guided by habits and that deeply ingrained habits can become “counter-intentional”, it then follows that people may often display “locked-in” practices in their daily energy consumption behaviour. This hypothesis has been assessed in our empirical analysis whose results show how the presence of strong energy-consuming habitual practices can reduce the effectiveness of economic incentives such as energy subsidies. One additional delicate factor that appears crucial for our purpose is that habits are not fully conscious forms of behaviours. This makes that individuals do not really see habits as a problem given that it is viewed as easily changed. In sum, based on our evolutionary account of the situation, it follows that, to be more efficient, climate policies would have to both shift the incumbent carbon-based socio-technical systems (for it to shape decisions towards a reduction of greenhouse gas emissions) and also deconstruct habits that this same socio-technical has forged with time (as increased environmental awareness and intentions formulated accordingly are not sufficient in the presence of strong habits). Accordingly, decision-makers should design measures (e.g. commitment strategies, niche management, etc.) that, as explained in this research, specifically target those change-resisting factors and their key features. This is essential as these factors tend to reduce the efficiency of traditional instruments. Micro-level interventions are thus needed as much as macro-level ones. For instance, it is often the case that external improvements of energy efficiency do not lead to lower energy consumption due to the rebound effect arising from unchanged energy-consuming habits. Bearing this in mind and building on the insights from the evolutionary approach, policy-makers should go beyond the mere subsidisation of technologies. They should instead create conditions enabling the use of the multi-layered, cumulative and self-reinforcing character of economic change highlighted by evolutionary analyses. This means supporting both social and physical technologies with the aim of influencing the selection environment so that only the low-carbon technologies and practices will survive. Mentioned references: David, P. A. (1985), Clio and the economics of QWERTY, American Economic Review 75/2: 332–337. IPCC, 2007, ‘Climate Change 2007: The Physical Science Basis’, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 996 pp. Stern, N., 2006, ‘Stern Review: The economics of Climate Change’, Report to the UK Prime Minister and Chancellor, London, 575 p. (www.sternreview.org.uk) van den Bergh, J.C.J.M., 2007, ‘Evolutionary thinking in environmental economics’, Journal of Evolutionary Economics 17(5): 521-549.

Du fait du développement de nouveaux produits (NPD) dans l’industrie, l’organisation devient de plus en plus complexe, ceci est dû notamment à la complexité même des produits. Dans ce contexte, le MBSE (Model Based Systems Engineering) et les approches collaboratives, qui adressent ces complexités, ont été reconnus pour leurs facultés à améliorer le NPD. Une implémentation réussie d’une conception collaborative du type MBSE, doit permettre de gérer ces deux complexités. Cette thèse de doctorat a pour objet l’étude de projets de conception collaborative MBSE au sein des équipes françaises chez des équipementiers automobiles et aéronautiques, afin de mettre en avant l’amélioration du développement des produits. La conception collaborative du type MBSE est assimilable à un système organisationnel complexe, impliquant des vues ou dimensions différentes. Ainsi, l’identification de ces dimensions, leur définition et l’étude de leurs interactions constituent le premier objectif de cette recherche. La compréhension de chacune d’entre elles pour améliorer la collaboration entre les différents membres du projet, est le deuxième objectif.Le troisième et dernier objectif de cette thèse est de proposer des systèmes socio-techniques (STS), assistant la collaboration. Les résultats de cette recherche, fournissent une méthodologie pour manager la complexité organisationnelle dans des projets collaboratifs du type MBSE. Elle est le produit d’une combinaison de quatre méthodes permettant la caractérisation de ses dimensions (processus, acteurs, objets et outils), tout en définissant leurs interactions. Ces méthodes assistent respectivement : 1) La description et l’évaluation de ces projets avec une perspective systémique 2) l’établissement d’une vision partagée du travail 3) l’analyse des coopérations entre les acteurs, et 4) le développement de STS tels quels des environnements collaboratifs et des supports collaboratif de capitalisation. L’implémentation en industrie des méthodes proposées, processus et recommandations, a montré comment la mise en avant de la collaboration dans les projets de design MBSE, permet d’améliorer l’ensemble du développement de produit. / The organisational complexity implied by New Product Development (NPD) within the industry, is often induced by the complex nature of the products themselves. In this context, MBSE (Model Based Systems Engineering) and collaborative approaches address those complexities and have been recognised by their contribution to improve the NPD processes. A successful implementation of a collaborative MBSE design would allow to manage both complexities. This PhD thesis describes an investigation on collaborative MBSE design projects within French teams in automotive and aeronautics companies, with the purpose of enhancing them to improve product development. We understand collaborative MBSE design as a complex organisational system which implies different views or dimensions. The identification of those dimensions, their definition and the study of their interactions constitute the first objective of this research. Understanding each dimension in order to improve collaboration between the project members is the second objectivedeuxième objectif.The third and last objective of this research is to propose Socio Technical Systems (STS) supporting this collaboration. The results of the thesis provide a methodology to manage organisational complexity in collaborative MBSE design projects. The methodology is a combination of four methods assisting the characterisation of the MBSE dimensions (people, process, information objects and tools), while defining their interactions. These methods support respectively: 1) The assessment and description of collaborative MBSE design projects from a systemic perspective 2) The establishment of a shared vision of the work 3) The analysis of the cooperation among the actors 4) The development of STS such as collaborative environment and a collaborative capitalisation support. The implementation of the proposed methods, process and guidelines in the industry has shown how the enhancement of collaboration in MBSE design projects can improve the overall product development.

Les plateformes numériques de travail permettent à des travailleurs et travailleuses de vendre leurs services en ligne. En croissance constante depuis les dernières années (Bureau international du travail, 2021 : 52), les caractéristiques de ce type d'activité s'étendent maintenant à d'autres types d'emplois salariés (Johanston et al., 2021 : 9). L'objectif de ce mémoire est de mieux comprendre le contrôle du travail dans ces relations triangulaires impliquant travailleurs et travailleuses, clientèle et plateforme. Plus précisément, nous cherchons à savoir comment et par quelle/s partie/s à la relation triangulaire s'exerce le contrôle du travail d'assistance virtuelle vendu sur une plateforme numérique de type marché. En utilisant une approche matérialiste, plus précisément la Labour Process Theory, nous avons documenté l'activité d'assistantes virtuelles identifiées comme canadiennes sur la plateforme Upwork, ici considérée comme un point de production digital. Cette étude de cas s'appuie sur un protocole de recherche mixte basé sur l'observation participante, l'analyse de contrats de travail, des entrevues semi-dirigées et des journaux de bord. Nos résultats, basés sur sept entrevues, 28 contrats et autant de journaux de bord, révèlent que l'assistance virtuelle sur Upwork est caractérisée par des mandats et des conditions de travail et d'emploi imprévisibles. Pour résoudre ces flous et performer sur ce marché, il est essentiel d'accomplir six modalités de travail invisible : de cumul, de recherche de contrats, de développement professionnel, de communication, de gestion de la réputation et émotionnel. Différentes ressources frontières, outils digitaux qui structurent la plateforme, seront utilisées par la clientèle et par Upwork afin de diriger, évaluer et discipliner (Kellogg et al. : 2020 : 380) les assistantes virtuelles. Est ainsi constitué un système sociotechnique, basé sur l'utilisation synergétique de ces ressources, dont la finalité est un autocontrôle de la travailleuse visant à assurer le respect des objectifs productifs de la clientèle et de la plateforme. La relation triangulaire est donc asymétrique et renouvèle les dynamiques d'exploitation présentes dans la relation salariale.

Le réchauffement climatique appelle, entre autres, à la réduction des émissions de gaz à effet de serre. Cette réduction passe par l'électrification du transport scolaire au Québec, une transition technologique. Il est problématique qu'elle progresse lentement. Geels affirme qu'elle est ralentie par l'action des acteurs dominants du secteur. Dans notre recherche, nous nous sommes demandé si les acteurs de l'autobus scolaire diesel résistent à l'arrivée de l'autobus scolaire électrique. Nos hypothèses postulaient qu'ils résistent. À l'aide de la perspective multiniveaux de Geels nous avons effectué 10 entrevues semi-dirigées avec des acteurs du secteur des autobus scolaires. Nos résultats corroborent seulement en partie nos hypothèses. Nos résultats indiquent aussi que ce sont davantage des facteurs de sélection, issues du gouvernement du Québec, qui apparaissent déterminants dans le succès de l'électrification de l'autobus scolaire. De futures recherches sur les transitions technologiques en transport auraient avantage à analyser les facteurs de sélection issus des actions des gouvernements, eux-mêmes poussés par l'opinion publique, deux facteurs qui se situent tous deux dans ce que Geels a appelé le paysage sociotechnique.

Le réchauffement climatique appelle, entre autres, à la réduction des émissions de gaz à effet de serre. Cette réduction passe par l'électrification du transport scolaire au Québec, une transition technologique. Il est problématique qu'elle progresse lentement. Geels affirme qu'elle est ralentie par l'action des acteurs dominants du secteur. Dans notre recherche, nous nous sommes demandé si les acteurs de l'autobus scolaire diesel résistent à l'arrivée de l'autobus scolaire électrique. Nos hypothèses postulaient qu'ils résistent. À l'aide de la perspective multiniveaux de Geels nous avons effectué 10 entrevues semi-dirigées avec des acteurs du secteur des autobus scolaires. Nos résultats corroborent seulement en partie nos hypothèses. Nos résultats indiquent aussi que ce sont davantage des facteurs de sélection, issues du gouvernement du Québec, qui apparaissent déterminants dans le succès de l'électrification de l'autobus scolaire. De futures recherches sur les transitions technologiques en transport auraient avantage à analyser les facteurs de sélection issus des actions des gouvernements, eux-mêmes poussés par l'opinion publique, deux facteurs qui se situent tous deux dans ce que Geels a appelé le paysage sociotechnique.

1. Contextual outline of the PhD Research<p><p>Climate change is today often seen as one of the most challenging issue that our civilisation will have to face during the 21st century. This is especially so now that the most recent scientific data have led to the conclusion that the globally averaged net effect of human activities since 1750 has been one of warming (IPCC 2007, p. 5) and that continued greenhouse gas emissions at or above current rates would cause further warming (IPCC, 2007 p. 13). This unequivocal link between climate change and anthropogenic activities requires an urgent, world-wide shift towards a low carbon economy (STERN 2006 p. iv) and coordinated policies and measures to manage this transition.<p><p>The climate issue is undoubtedly a typical policy question and as such, is considered amenable to economic scrutiny. Indeed, in today’s world economics is inevitable when it comes to arbitrages in the field of policy making. From the very beginning of international talks on climate change, up until the most recent discussions on a post-Kyoto international framework, economic arguments have turned out to be crucial elements of the analysis that shapes policy responses to the climate threat. This can be illustrated by the prominent role that economics has played in the different analyses produced by the Intergovernmental Panel on Climate Change (IPCC) to assess the impact of climate change on society.<p><p>The starting point and the core idea of this PhD research is the long-held observation that the threat of climate change calls for a change of climate in economics. Borrowing from the jargon used in climate policy, adaptation measures could also usefully target the academic discipline of economics. Given that inherent characteristics of the climate problem (e.g. complexity, irreversibility, deep uncertainty, etc.) challenge core economic assumptions, mainstream economic theory does not appear as appropriately equipped to deal with this crucial issue. This makes that new assumptions and analyses are needed in economics in order to comprehend and respond to the problem of climate change.<p><p>In parallel (and without environmental considerations being specifically the driving force to it), the mainstream model in economics has also long been (and still is) strongly criticised and disputed by numerous scholars - both from within and outside the field of economics. For the sake of functionality, these criticisms - whether they relate to theoretical inconsistencies or are empirically-based - can be subsumed as all challenging part of the Cartesian/Newtonian legacy of economics. This legacy can be shown to have led to a model imprinted with what could be called “mechanistic reductionism”. The mechanistic side refers to the Homo oeconomicus construct while reductionism refers to the quest for micro-foundations materialised with the representative agent hypothesis. These two hypotheses constitute, together with the conjecture of perfect markets, the building blocks of the framework of general equilibrium economics. <p><p>Even though it is functional for the purpose of this work to present them separately, the flaws of economics in dealing with the specificities of the climate issue are not considered independent from the fundamental objections made to the theoretical framework of mainstream economics. The former only make the latter seem more pregnant while the current failure of traditional climate policies informed by mainstream economics render the need for complementary approaches more urgent. <p><p>2. Overview of the approach and its main insights for climate policy<p><p>Starting from this observation, the main objective of this PhD is thus to assess the implications for climate policy that arise from adopting an alternative analytical economic framework. The stance is that the coupling of insights from the framework of evolutionary economics with the perspective of ecological economics provides a promising way forward both theoretically as well as on a more applied basis with respect to a better comprehension of the socioeconomic aspects related to the climate problem. As claimed in van den Bergh (2007, p. 521), ecological economics and evolutionary economics “share many characteristics and can be combined in a fruitful way" - which renders the coupling approach both legitimate and promising. <p><p>The choice of an evolutionary line of thought initially stems from its core characteristic: given its focus on innovation and system change it provides a useful approach to start with for assessing and managing the needed transition towards a low carbon economy. Besides, its shift of focus towards a better understanding of economic dynamics together with its departure from the perfect rationality hypothesis renders evolutionary economics a suitable theoretical complement for designing environmental policies.<p><p>The notions of path-dependence and lock-in can be seen as the core elements from this PhD research. They arise from adopting a framework which is founded on a different view of individual rationality and that allows for richer and more complex causalities to be accounted for. In a quest for surmounting the above-mentioned problem of reductionism, our framework builds on the idea of ‘multi-level selection’. This means that our analytical framework should be able to accommodate not only for upward but also for downward causation, without giving analytical priority to any level over the other. One crucial implication of such a framework is that the notion of circularity becomes the core dynamic, highlighting the importance of historicity, feedbacks and emergent properties. <p><p>More precisely, the added value of the perspective adopted in this PhD research is that it highlights the role played by inertia and path-dependence. Obviously, it is essential to have a good understanding of the underlying causes of that inertia prior to devising on how to enforce a change. Providing a clear picture of the socio-economic processes at play in shaping socio-technical systems is thus a necessary first step in order to usefully complement policy-making in the field of energy and climate change. In providing an analytical basis for this important diagnosis to be performed, the use of the evolutionary framework sheds a new light on the transition towards low-carbon socio-technical systems. The objective is to suggest strategies that could prove efficient in triggering the needed transition such as it has been the case in past “lock-in” stories. <p><p>Most notably, the evolutionary framework allows us to depict the presence of two sources of inertia (i.e at the levels of individuals through “habits” and at the level of socio-technical systems) that mutually reinforce each other in a path-dependent manner. Within the broad perspective on path dependence and lock-in, this PhD research has first sketched the implications for climate policy of applying the concept of ‘technological lock-in’ in a systemic perspective. We then investigated in more details the notion of habits. This is important as the ‘behavioural’ part of the lock-in process, although explicitly acknowledged in the pioneer work of Paul David (David, 1985, p. 336), has been neglected in most of subsequent analyses. Throughout this study, the notion of habits has been studied at both the theoretical and applied level of analysis as well as from an empirical perspective. <p><p>As shown in the first chapters of the PhD, the advantage of our approach is that it can incorporate theories that so far have been presented opposite, partial and incomplete perspectives. For instance, it is shown that our evolutionary approach not only is able to provide explanation to some of the puzzling questions in economics (e.g. the problem of strong reciprocity displayed by individual in anonymous one-shot situations) but also is very helpful in bringing a complementary explanation with respect to the famous debate on the ‘no-regret’ emission reduction potential which agitates the experts of climate policy. <p><p>An emission reduction potential is said to be "no regret" when the costs of implementing a measure are more than offset by the benefits it generates such as, for instance, reduced energy bills. In explaining why individuals do not spontaneously implement those highly profitable energy-efficient investments ,it appears that most prior analyses have neglected the importance of non-economic obstacle. They are often referred to as “barriers” and partly relate to the ‘bounded rationality’ of economic agent. As developed in the different chapters of this PhD research, the framework of evolutionary economics is very useful in that it is able to provide a two-fold account (i.e. relying on both individual and socio-technical sources of inertia) of this limited rationality that prevent individuals to act as purely optimising agents.<p><p>Bearing this context in mind, the concept of habits, as defined and developed in this study, is essential in analysing the determinants of energy consumption. Indeed, this concept sheds an insightful light on the puzzling question of why energy consumption keeps rising even though there is an evident increase of awareness and concern about energy-related environmental issues such as climate change. Indeed, if we subscribe to the idea that energy-consuming behaviours are often guided by habits and that deeply ingrained habits can become “counter-intentional”, it then follows that people may often display “locked-in” practices in their daily energy consumption behaviour. This hypothesis has been assessed in our empirical analysis whose results show how the presence of strong energy-consuming habitual practices can reduce the effectiveness of economic incentives such as energy subsidies. One additional delicate factor that appears crucial for our purpose is that habits are not fully conscious forms of behaviours. This makes that individuals do not really see habits as a problem given that it is viewed as easily changed.<p><p>In sum, based on our evolutionary account of the situation, it follows that, to be more efficient, climate policies would have to both shift the incumbent carbon-based socio-technical systems (for it to shape decisions towards a reduction of greenhouse gas emissions) and also deconstruct habits that this same socio-technical has forged with time (as increased environmental awareness and intentions formulated accordingly are not sufficient in the presence of strong habits).<p><p>Accordingly, decision-makers should design measures (e.g. commitment strategies, niche management, etc.) that, as explained in this research, specifically target those change-resisting factors and their key features. This is essential as these factors tend to reduce the efficiency of traditional instruments. Micro-level interventions are thus needed as much as macro-level ones. For instance, it is often the case that external improvements of energy efficiency do not lead to lower energy consumption due to the rebound effect arising from unchanged energy-consuming habits. Bearing this in mind and building on the insights from the evolutionary approach, policy-makers should go beyond the mere subsidisation of technologies. They should instead create conditions enabling the use of the multi-layered, cumulative and self-reinforcing character of economic change highlighted by evolutionary analyses. This means supporting both social and physical technologies with the aim of influencing the selection environment so that only the low-carbon technologies and practices will survive. <p><p><p>Mentioned references:<p><p>David, P. A. (1985), Clio and the economics of QWERTY, American Economic Review 75/2: 332–337.<p><p>IPCC, 2007, ‘Climate Change 2007: The Physical Science Basis’, Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S. D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M. Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 996 pp. <p><p>Stern, N. 2006, ‘Stern Review: The economics of Climate Change’, Report to the UK Prime Minister and Chancellor, London, 575 p. (www.sternreview.org.uk)<p><p>van den Bergh, J.C.J.M. 2007, ‘Evolutionary thinking in environmental economics’, Journal of Evolutionary Economics 17(5): 521-549.<p> / Doctorat en Sciences économiques et de gestion / info:eu-repo/semantics/nonPublished