Thèse Evaluer l'Étendue et Explorer les Mécanismes de l'Effet-Rebond Circulaire dans l'Économie de Seconde Main H/F - 75

Établissement : Université Paris-Saclay GS Economie & Management
École doctorale : Droit, Economie, Management
Laboratoire de recherche : CEPS - Centre d'Économie de l'ENS Paris-Saclay
Direction de la thèse : Vincent MARTINET ORCID 0000000271358849
Début de la thèse : 2026-09-01
Date limite de candidature : 2026-04-15T23:59:59

L'économie circulaire est promue depuis plusieurs décennies pour ses vertus écologiques, tant dans les pratiques militantes qu'au niveau des politiques publiques. Cette forme d'organisation économique vise à prolonger les possibilités d'usage des biens afin d'accroître l'utilité tirée des matériaux et de l'énergie nécessaires à leur production. Elle englobe un large éventail de pratiques telles que la réparation, le recyclage, le reconditionnement et la consommation de biens d'occasion (Zink et al., 2017). Cette dernière, en particulier, a connu une croissance substantielle ces dernières années, notamment en raison de la réduction drastique des coûts de transaction permise par l'émergence de plateformes en ligne dédiées (Dekhili et al., 2025). Si l'avantage environnemental de la production secondaire est indéniable d'un point de vue strictement physique - puisqu'elle permet d'économiser l'extraction de nouvelles ressources - la majorité des études sur le sujet ne prennent pas en compte les mécanismes de marché qui la structurent. Ces mécanismes peuvent engendrer un « effet rebond » qui réduit, voire annule, les bénéfices environnementaux attendus. Au cours de la dernière décennie, la recherche en économie s'est de plus en plus saisie de ces questions, tant sur le plan théorique qu'empirique, mais elle ne dispose pas encore d'un cadre analytique consensuel et unifié. Cette thèse vise donc à contribuer à ce champ de recherche émergent en se concentrant sur l'économie de l'occasion.

Since the 1970s, environmental scientists have advocated for the circular economy, and it is now commonly accepted that such practices allow for savings in energy and manufacturing materials while reducing the environmental impact of consumption and production. Graedel and Allenby (1996) even argue that the ultimate goal of an economic organization compatible with environ- mental limits would be an economic system in which all materials are reused or recycled. The circular economy can operate at three levels: material recovery (recycling), component recovery (remanufacturing), and product-level recovery (repair, refurbishment, or second-hand use). Products resulting from these practices are grouped under the term secondary production (Zink and Geyer 2017). Research efforts in environmental sciences have therefore mainly focused on evaluating differences in environmental footprints between goods originating from primary production and those from secondary production, as well as on developing rigorous measurement methodologies such as life cycle assessment (e.g., Atherton 2007; Ekvall and Finnveden 2001; Weidema 2003). From a policy perspective, the circular economy is now strongly supported by governments and consumers alike, and it is frequently presented as a marketing argument by firms involved in these practices. The European Commission, for example, allocated a €6 billion fund in 2016 to foster the development of the circular economy (EC 2016). Within this circular economy, the second-hand market has experienced particularly strong growth, especially in the fashion sector. In France alone, the secondhand fashion market was estimated at €1 billion in 2022 (KPMG and FEVAD, 2022) and could even surpass the primary fashion market by 2028 (IndexPresse Etudes, 2020). This growth is mainly explained by the emergence of thrift stores, but above all by peer-to-peer digital platforms specialized in fashion, such as the French platforms 'Videdressing' and 'Vestiaire Collective', and especially the Lithuanian platform 'Vinted', valued at over €3.5 billion in 2021(Levingston and Milne, 2023), for which France represents the largest European market (Simon Rainaud, 2023). Second-hand consumption has thus become mainstream, with 54% of French consumers reporting participation in second-hand markets (Ademe, 2022). Far from being limited to fashion, the second-hand economy represents a substantial share of exchanges across a wide range of markets (equipment, automobiles, household appliances, etc.). Platforms emphasize the environmental dimension of these practices, which extend product lifespans, and this is often cited by consumers as one of their main motivations (Kurisu et al., 2021[24]). However, these analyses often assume that substitution between a good from primary production and a good from secondary production occurs on a one-to-one basis (Graedel and Allenby 1996; Curran 1996; Stahel 1994). Under this assumption, as formalized by Zink and Geyer (2017)[39], the
environmental benefit of the circular economy can be expressed as: Enet = (er ep)Qr where er denotes the environmental impact of reuse, ep the environmental impact of producing a new good, and Qr the quantity of reused goods. This formulation abstracts from the market mechanisms governing production and consumption by assuming fixed total consumption, thereby implying an automatic displacement of new goods by reused goods. In reality, the environmental impact of the development of second-hand markets should be measured as: Enet = erQr +epQp which corresponds to the previous equation once the assumption Qr = Qp is relaxed. In practice, it is highly likely that the production and consumption of reused goods do not reduce-and may even
increase-the production and consumption of new goods. This phenomenon is referred to as the rebound effect. The rebound effect is an old concept originating in the energy economics literature. It was first developed by Stanley Jevons in 1865 following his observations of the coal market, which led him to highlight the coexistence of substantial gains in energy efficiency on the one hand and rising coal consumption on the other. The concept re-emerged in the 1970s alongside early efforts to regulate energy consumption. The works of Khazzoom (1980) and Brookes (1990), in particular, played a major role in reviving and formalizing the concept. Economics has since developed a large body of literature describing this phenomenon. Research on the rebound effect has covered a
wide range of sectors, including transportation (de Haan et al. 2007; Greening et al. 2000; Spielmann et al. 2008), lighting (Fouquet and Pearson 2012; Saunders and Tsao 2012), and even non-energy goods (Chalmers et al. 2015; Thiesen et al. 2008). A literature review by Greening et al. (2000) establishes a typology of rebound effects according to their underlying mechanisms: First, there are direct rebound effects, whereby energy efficiency gains lead to lower prices of energy-intensive goods, either in production or in use, generating increased consumption of these goods and, subsequently, higher energy use. Next are secondary effects, which refer to cases in which energy savings in one sector made more efficient allow for an increase in energy demand in other sectors.
The third category consists of economy-wide effects, which are similar to secondary effects but broader and less predictable; they refer to overall economic growth enabled by energy efficiency gains, which once again increases energy demand. Finally, transformational effects encompass rebound effects arising from broader changes in consumer preferences, institutions, or regulation induced by energy efficiency improvements. While rebound effects were initially studied mainly in relation to efficiency gains on the production side, the analysis has since been extended to efficiency gains on the consumer side, notably those enabled by improved usage behaviors (Borenstein 2013). Given the conceptual proximity between rebound effects in energy markets driven by energy
efficiency gains and rebound effects in primary goods markets driven by increases in utility derived from goods through second-hand markets or the circular economy more generally, Zink and Geyer (2017) introduce the concept of Circular Economy Rebound. Their explicit objective is to standardize the terminology used in the economic literature in order to facilitate the development of theoretical and empirical research aimed at characterizing this phenomenon. They identify at least two potential sources of circular economy rebound effects. The first arises from income effects generated by second-hand markets and by the circular economy more broadly. Through more intensive use of goods, their components, or their materials, the circular economy lowers prices relative to the
utility derived from consumption. For instance, second-hand goods are inexpensive for secondary consumers while simultaneously providing income to primary consumers, effectively acting as if the primary good were cheaper. This generates an income effect that can expand the market and strengthen demand for the good in question or for other goods more generally. The second source of circular economy rebound effects stems from imperfect substitution between secondary goods and primary goods. Some recycled materials are of lower quality than their primary-market equivalents, making them imperfect substitutes. This is the case for aluminum (Nakajima et al. 2010), paper, and plastics (Allwood 2014). Similarly, second-hand goods may not satisfy the same needs or align
with the same preference profiles as their new counterparts. This particularly concerns goods for which obsolescence relative to new products is rapid or reinforced by fashion effects. Second-hand smartphones, for example, compete only weakly with new devices due to their real or perceived obsolescence among consumers with a strong preference for novelty. In practice, they are mainly sold in developing countries where new smartphones are prohibitively expensive (Geyer and Doctori Blass 2010; Skerlos et al. 2003). In such cases, the circular economy adds a new market alongside the primary market rather than replacing it, thereby very likely increasing total production. As noted by Zink and Geyer (2017), there is a considerable body of research quantifying the environmental impacts of many different recycling, refurbishment, and repair activities. In stark contrast, the impact of those activities on primary material and product production is largely unknown. Nonetheless, several indicators suggest that rebound effects should be taken very seriously. Zink et al. (2016) show
that perfect substitution of recycled materials for primary materials occurs only under specific parametric conditions that are unlikely to hold in competitive markets. With regard to second-hand markets in particular, few theoretical papers address this question. The earliest contribution dates back to Fox (1957), who provides a graphical analysis of the interactions between primary and secondary markets. He shows that second-hand markets may reinforce primary markets by transforming goods owned by primary consumers into liquid assets that can be easily resold, thereby facilitating faster replacement with identical new goods. The work of Scitovsky (1994) points in a similar direction. Kürsten (1991) develops a model showing how the resale of durable goods, such as automobiles, supports aggregate demand through an income effect by increasing consumers' overall wealth. Finally, Thomas (2003)[36] theoretically demonstrates an ambiguous effect of second-hand markets on primary markets. Building on a model by Anderson and Ginsburgh (1994)[5], he shows that a
reduction in transaction costs on second-hand markets allows consumers who were previously excluded from the market-and even some former primary buyers-to purchase goods on the secondary market at a price equal to transaction costs. This expansion of the secondary market therefore leads to a contraction of the primary market, although to a lesser extent due to the entry of new consumers previously excluded from consumption. However, once transaction costs become sufficiently low, the quantity of goods available on the second-hand market becomes smaller than demand at a price equal to transaction costs. Second hand prices then increase beyond transaction costs, allowing primary buyers to earn resale profits, which effectively lowers the net cost of first-hand consumption and stimulates demand in the primary market. From an empirical perspective, the most active research on this topic mainly comes from management science and marketing, and studies the effect of second-hand categorization on consumer behavior. Based on data from the Vinted platform and around forty interviews, Dekhili et al. (2025) show that online second-hand platforms encourage overconsumption through the promise of a renewed consumption (or selling) experience and ease of purchase, which reinforces impulsive buying behavior. Consumers, for their part, value the environmentally friendly aspect of these practices, which often allows them to justify purchases that they sometimes criticize when they are new, such as fast fashion. Finally, they observe that second-hand purchases often coexist with purchases of new goods within the practices of the same consumer, indicating imperfect substitutability that is conducive to a rebound effect. The issue of rebound effects in second-hand markets therefore deserves to be studied from both a theoretical and an empirical perspective. However, the authors discussed above themselves regret the lack of a consensual theoretical framework for second-hand markets and their economic impacts (Thomas, 2003). The models used so far are ad hoc models that make it possible to highlight the relevance of searching for a rebound effect in second-hand markets, but not to understand its concrete mechanisms and conditions of emergence. In particular, these models lack a micro founded explanation for the existence of a secondary market alongside a primary market, that is, an explanation of what may lead two consumers to value the same good differently, prompting one to acquire it and the other to dispose of it. The model proposed by Thomas (2003), for example, imposes the existence of a second-hand market through the assumption that first-hand buyers part with the good at the end of a period regardless of circumstances, either by reselling it or discarding it, even though its lifetime is 1 + L periods (L < 1). The allocation between first-hand buyers, second-hand buyers, and non-buyers is thus determined by heterogeneous preferences for the product, leading the former to purchase it at a higher price for a usage period of length 1, while the latter prefer to own it for a shorter period (L) but at a lower price. The model of Fox (1957) instead relies on product degradation, which imposes increasingly high usage costs that differ across consumers, some of whom possess greater skills, time, or even an intrinsic interest in maintenance and refurbishment activities. This leads to consumer-specific depreciation rates for the good. This justification fits well with durable goods markets such as automobiles, which were particularly studied in the previous century. Following Akerlof (1970) and his theory of adverse selection in second-hand car markets due to information asymmetries, a number of studies have demonstrated this phenomenon in other durable goods markets (Rothschild and Stiglitz, 1976; Bond, 1982; Van Cayseele, 1993; Hendel and Lizzeri, 1999). Waldman (2003)[38], in particular, provides a comprehensive literature review. However, this logic of differentiated depreciation rates as an explanation for the existence of second-hand markets for durable goods, whose quality visibly deteriorates between initial purchase and resale, does not fit well with other rapidly growing second-hand markets today, such as clothing or smartphones. It is indeed questionable to assume that the existence of second-hand fashion markets is due solely to differences in consumers' ability to sew or store garments whose quality has degraded, or even to differences in quality valuation. The models examined so far also suffer from the absence of an alternative consumption option alongside the primary and secondary markets (other than the null option of not purchasing anything), which prevents distinguishing indirect rebound effects from direct ones. Many of them, for example, impose that any good sold on the secondary market is replaced by a purchase on the primary market (Fox, 1957; Scitovsky, 1994), thereby effectively denying any potential ecological benefit of second-hand consumption. Finally, new and second-hand goods are often treated as homogeneous except for quality, which contradicts empirical studies showing imperfect substitutability (Geyer and Doctori Blass, 2010; Skerlos et al., 2003).

Etudier l'effet de l'existence d'un marché de seconde-main sur les impacts environnementaux de la consommation, et en particulier la possibilité d'effets rebond.