Abstract
We conceptualize a globally integrated circular economy through the notion of global circular networks (GCN) and identify the key mechanisms of circular (de-)coupling. We develop a novel conceptual framework by integrating concepts from literatures on global production networks (GPN) and the circular economy (CE). The framework combines four essential dimensions of these couplings—actors, resources, functions, and geography—which we illustrate with empirical examples. Our framework establishes a multi-scalar understanding of the CE that can inform both empirical research approaches and policy making at different levels on how to rethink global entanglements and drive circularity-oriented changes.
Introduction
The global production of goods is a major contributor to climate change, environmental degradation, resource depletion, and uneven global development (e.g., Meng et al., 2018; Ponte et al., 2023). To a large extent, production is organized in global production networks (GPNs, e.g., Coe and Yeung, 2015). These constitute the architecture for current linear material flows that underpins the dominating take-make-waste logic across space and scale (Ellen Macarthur Foundation, 2023). How these networks can be transformed to become more sustainable is a key question. We argue that this would fundamentally challenge their linear logic, especially their resource flows 1 in production and consumption processes, thus relating to aspects such as material footprints, throughput, and emissions as well as their changing geography. In this regard, the circular economy (CE)—as opposed to the linear economy—is increasingly seen as a desirable alternative to drive sustainable change. It represents an alternative model for economic processes and organization, represented in the circulation of resources as a driving metaphor (Geissdoerfer et al., 2017; Korhonen et al., 2018). The concept is prominently favored in policy strategies and academic research to successfully combine the goals of economic growth and environmental sustainability. Furthermore, multinational companies (MNCs) 2 are increasingly articulating strategies for circular orientation of their global operations and processes, which could eventually affect the geography of their operations. The implementation of the CE varies across industries. For example, the paper and pulp industry has well established national systems for collecting and recycling newspapers and packaging. Nonetheless, nearly a fifth of the material is still exported for recycling (Provost-Savard et al., 2023). The fashion industry, by contrast, is only beginning to develop textile-to-textile recycling systems, often located far from original production sites. This will require the industry to develop new functions (collection, sorting, recycling), some of which need to be located geographically close to consumers (see e.g., Bryson et al., 2024).
Against this backdrop, the question arises as to how circular changes in GPNs, currently dominated by the linear paradigm, can be theorized and situated within geographical thinking. Two key research fields are relevant for this: research on GPNs (e.g., Coe et al., 2004; Coe and Yeung, 2015; Henderson et al., 2002) and the CE (e.g., Geissdoerfer et al., 2017; Korhonen et al., 2018). However, both research on the CE and within the GPN literature lack a conceptual understanding and integration of how circular changes can unfold in these global networks. Moreover, due to opposing logics (linear versus circular), and different implied geographies, it is not evident if, how, when, and where real-world changes could be possible. On the one hand, research on GPNs has been silent on issues related to sustainability, including the CE (Coe and Gibson, 2023). While this could be the result of a lack of empirical cases from which to theorize, it also points to conceptual blind spots and ontological traditions. On the other hand, research on the CE largely lacks a spatial perspective; one that understands the geography of circular changes, how this relates to globally integrated economic processes, and the multi-scalar architecture of circular trajectories. Among the rather few geographical contributions to the vast and quickly expanding CE literature (e.g., Bourdin and Torre, 2024; Davies et al., 2024; Deutz et al., 2024; Hobson, 2015; Meili et al., 2024; Rülke et al., 2025), a global perspective that acknowledges the significance of the functional integration of economic activities in GPNs is largely missing. Rather, the CE is often understood as a locally or regionally oriented form of economic organization. Yet, geographical contributions have long underscored the global (and unequal) nature of waste streams and treatment (e.g., Crang et al., 2012; Gregson and Crang, 2015; Gregson et al., 2011).
Against this background, in this paper, we argue that considerations of the CE must take into account the existing global architecture of the economy in GPNs as the basis for structural reorientation and economic evolution. While we can perceive a rescaling of the economy in space and time, where firm activities and relationships may move from being fully global to more regional (Grillitsch et al., 2025a), we find it difficult to imagine an absolutely deglobalized world, given the many everyday aspects of globalization, transnational infrastructures, and the path dependent nature of industrial development. Therefore, there is a critical need to conceptualize and develop a deep understanding of the interrelationships between the CE and the global economic architecture to unlock circular potentials. We pursue this novel conceptual endeavor, cultivated in an in-depth discussion and problematization of the assumptions (cf. Alvesson and Sandberg, 2011) that characterize literatures on GPNs and the CE, and that relate to differences in both epistemological origins and ontological traditions. We integrate key assumptions from both literatures and develop a framework of global circular networks rooted in a functionalist understanding of the geography of these networks. We furthermore conceptualize circular (de-)couplings as a necessary, but not sufficient, mechanism for a shift from the global linear economy to the global CE. This mechanism captures the transformation of GPNs into global circular networks (GCN), based on couplings with circular functions (i.e., techno-economic capabilities, see section “Circular (de-)couplings and functions”). Hence, this paper proposes a theorization of circular transitions, starting from GPNs situated in the linear paradigm and within a multi-scalar architecture, along with a conceptual framework that provides analytical purchase for empirical exploration. This way, we complement existing GPN and CE research while also moving beyond it by presenting a dynamic view of the shift from GPNs to GCNs. We pose the following research questions: • How can a globally integrated circular economy be conceptualized? • What are the mechanisms to transition from the linear organization of global production networks to a (globally integrated) circular economy?
The paper is organized as follows. We first describe the assumptions that characterize the GPN and CE research fields, before conceptualizing a globally integrated economy through the notion of GCNs and the central process of circular (de)couplings. We close with a discussion of the emerging field of geography of the CE and a brief conclusion.
Assumptions and limitations of GPNs and CE as research fields
The discourse and research on the CE pay little attention to the global integration of economic activities, which is a major shortcoming we address by drawing on literature about GPNs. Within the social sciences, the global economy is commonly examined through the literatures on GPNs and Global Value Chains (GVCs), which, while overlapping in scope, diverge in their conceptual emphasis. In this paper, we focus on GPN scholarship since it is more sensitive to the spatial and multi-scalar dimensions of value creation and governance, in part due to the reliance on a linear chain metaphor within GVC literature. Hence, we argue that the GPN lens offers greater potential for identifying and problematizing the globally integrated dynamics of the CE. Especially the concept of couplings (e.g., Coe and Yeung, 2015) enables the analysis of geographically sensitive changes in the development of a globally integrated CE, allowing for a deeper exploration of why, how, where, and when production networks are being reconfigured for enhanced circulation of resource flows. Other research fields, such as transition studies, make important contributions to understanding the socio-technical changes involved in circular transitions, some of which we outline in the subsection on assumptions of the CE literature. Yet, transition research has paid less attention to the actors that reproduce, in the institutional sense, the core global regime of production-oriented sectors and produce and distribute, in the manufacturing sense, goods from unsustainable resource flows. With our theoretical situation in GPN scholarship, we therefore complement a stream of research that calls for more attention to the role of MNCs and other globally hegemonic actors in (circular) transitions, and their organization in GPNs (cf., Fuenfschilling and Binz, 2018; Lesch et al., 2023).
Against this backdrop, we will unpack the assumptions and limitations within existing research on GPNs and the CE using a problematization approach (Alvesson and Sandberg, 2011), before explaining the processes underlying the formation of a globally integrated CE in the next section. Using this approach, we identify and challenge the underlying assumptions of concepts and theories (as an alternative to a gap spotting approach) in the two streams of literature. We operationalized the approach through reading in-depth key texts 3 in both literatures, followed by intensive discussions among the authors of this paper. The readings were chosen to cover key texts and recent critiques to provide a comprehensive overview of both literatures and were complemented with additional readings. The following two subsections summarize these discussions.
Assumptions characterizing the GPN literature
When the concept of GPNs was developed by Henderson et al. (2002), it was characterized by carefully chosen considerations of using the word “global” instead of “international” and “transnational” to break with state-centric discourses. It is considered that the functional integration of activities in GPNs, following globalization, reflects a qualitatively different phenomenon to previous modes of production. The concept focuses on immaterial flows (value, knowledge, and power) in multi-actor settings, with actors driven by different motivations. The metaphor of networks is central as it embraces vertical and horizontal integration. It is rooted in multi-scalarity (vertical and horizontal relationships between places), a linear economy (one-flow direction of materials, goods, and products) (Herod et al., 2013), and relationality (actors need to be considered in relation to each other). Furthermore, it is shaped by the idea that globalization is a conducive condition for economic growth and prosperity, while not considering material resources as a constraint. Consequently, a driving assumption that emerged from early work is that understanding globalization (and especially post-Fordism) requires a shift in the analysis of the economy towards the global processes and architecture of economic organization (Henderson et al., 2002). Coe et al. (2004) operationalize this by focusing on MNCs (lead firms) as central actors and assume that strategic coupling is a necessary condition for regional development. This work perceives MNCs as superior to other firms, states, and regions, and of being flexible in their location decisions. Power relations at the state-firm nexus, between MNCs and regions, and between MNCs and other firms have become a central aspect in the analysis of global relations, while couplings have become a central mechanism. Power is considered to lie mainly with lead firms, while states provide and constitute institutional environments.
The normative orientation of GPN research is to promote economic growth in regions and states, which could be achieved through the three main mechanisms of value creation, enhancement, and capture, as well as the exploitation of scale and scope economies. GPN research assumes a global competition between regions and states in order to attract MNCs (Coe et al., 2004). Yeung and Coe (2014) introduced what they call GPN 2.0, a reconceptualization of GPNs to respond to criticisms regarding GPN being a static approach, which fails to account for development dynamics. Inherent in this dynamic view on GPNs is the assumption that development outcomes depend on GPNs, which in turn evolve as a result of decisions of rational economic actors bargaining in markets and their competitive dynamics. This underscores that GPNs are primarily understood through lead firms and from a micro-perspective of their actions and interactions (Yeung and Coe, 2014).
Couplings, meaning the integration of regions in GPNs, are nuanced in this dynamic view by differentiating between structural, functional, and organic couplings. Structural couplings refer to situations in which “external actors connect the region into [GPNs]” (Coe and Young, 2015, p. 184). These are relatively stable and deeply integrated forms of couplings, for example, through the presence of a MNC in the region or Foreign Direct Investment. Organic couplings describe situations in which “regional actors reach outside their home region to construct [GPNs]” (p. 184). Here, the initiative comes from regional actors, which results in a strong integration in the GPN. Functional couplings describe the process by which “regional actors productively meet the wider needs of [GPNs]” (Coe and Yeung, 2015: 184). This results in moderate integration within GPNs, which can be enhanced through upgrading. Particularly functional couplings may also be conveyed through intermediaries that organize knowledge and information flows between MNCs and local firms (such as between Asian garment manufacturers and European fashion labels) (Coe and Yeung, 2015). The more integrated actors are in GPNs, the more stable these couplings are assumed to be, leading actors to compete for upgrading. Couplings may be shaped by friction and asymmetric power relations, for instance between small-scale producers seeking opportunities to upgrade within GPNs and MNCs aiming to capture the largest share of value from their suppliers. Recent contributions have further developed this notion discussing decoupling from regional economies due to geopolitical change and instability (Blažek and Lypianin, 2024; Pavlínek, 2023). Nowadays, these different forms of couplings that shape the functionally integrated global economy as well as how they are negotiated at the state-firm nexus, are considered a central mechanism and analytical category of GPN research. This has led to the assumption that regions and states need to exploit their techno-economic capabilities and institutional environments to link up to GPNs by means of strategic couplings, and through this, foster economic development. Based on this, the notion that GPNs are critical to regional development has matured and become a field assumption within economic geography and beyond.
However, the GPN concept has also faced considerable criticism, and a variety of alternative perspectives and extensions have emerged (e.g., Bair and Werner, 2011; Crang et al., 2012; Gregson and Crang, 2015; Ibert et al., 2019; Kleibert et al., 2020; MacKinnon, 2012; McGrath, 2017; Werner, 2016; Werner, 2018; Werner, 2020). Among these, MacKinnon (2012) criticizes the neglect of uneven power relations, path dependency and lock-ins, which create specific spatio-temporal configurations of possibilities for regional development. MacKinnon (ibid.) argues that it is important to consider regional differences which are situated in an evolutionary paradigm to account for the organic, rather than purely strategic, nature of couplings. He furthermore argues that recouplings and decouplings in GPNs can be understood as drivers of uneven development resulting from socioeconomic losses when MNCs may leave regions.
While other authors have raised issues related to the inclusionary bias (Werner, 2016) of GPNs and the lack of engagement with dissociations (Ibert et al., 2019), we have observed a general lack of engagement with the directionality of GPNs apart from a focus on economic development and growth. This is prevalent in a lack of studies that use a GPN perspective to understand sustainability-oriented shifts in the global architecture of the economy. Consequently, there is a severe lack of knowledge of how directed (in a normative sense) processes can unfold in a multi-scalar architecture and what their central mechanisms are. This is unfortunate with ongoing environmental and climate change, and asks for novel contributions that are able to conceptually embrace sustainability- or circular-oriented reconfigurations in the global economy that might involve rescaling of economic relationships (Grillitsch et al., 2025a). We will return to this issue in the next section.
Assumptions characterizing the CE literature
Although the CE as both a concept and empirical phenomena has recently attracted the attention of scholars from different disciplines with a booming number of articles (Calisto Friant et al., 2020; Kirchherr et al., 2017), the concept’s core message and metaphors of closed loops and circles date back to the 1970s and 1980s (Geissdoerfer et al., 2017; Pearce and Turner, 1989; Stahel and Reday, 1976). CE research focuses on material flows and proposes a regenerative and restorative system of production (Calisto Friant et al., 2020; Geissdoerfer et al., 2017). Within a CE, power is considered to mainly lie with governments and companies (Geissdoerfer et al., 2017), and the idea of the CE has been strongly developed within these groups. Among others the EU (e.g., European Circular Economic package of 2015) and individual states within Europe (e.g., Germany with its Closed Substance Cycle and Waste Management Act of 1996) as well as China (e.g., Circular Economy Promotion Law of 2008) have adopted the concept into policies. This also influences how the concept is used and understood. CE is often perceived as a toolbox to address sustainable development (Leipold et al., 2022) and is “expected to solve problems of resource scarcity, biochemical flow disruption, and climate change, all while revitalizing local and regional economies” (Calisto Friant et al., 2020: 1). A focus on economic prosperity is particularly common in definitions used by practitioners (Kirchherr et al., 2017), and the beneficiaries of the CE are seen as the economic actors who implement the system (Geissdoerfer et al., 2017). A CE is expected to extend the life of products through, for example, services, employment is expected to be positively influenced in places of consumption (such as Europe), while places of production (such as Asia) may be negatively affected by both reduced sales and employment (Skene, 2017). The idea that the understanding of CE is largely practitioner-driven is a common assumption and is an important backdrop when understanding how scholars perceive and problematize the concept.
The metaphors of closed and open loops (or loop economy) and circles are central in the scholarly CE literature (Calisto Friant et al., 2020; Corvellec et al., 2021; Geissdoerfer et al., 2017; Korhonen et al., 2018). So is the waste hierarchy and the R’s framework, outlining different R’s such as Reduce, Reuse, Recycle, and Recover (Kirchherr et al., 2017; Korhonen et al., 2018; Reike et al., 2018). However, the CE concept is also ambiguous and encompasses different meanings. Some scholars see this in a rather positive light (Leipold et al., 2022), in the sense that it is adaptable and promotes creativity among practitioners. Others instead perceive it as negative, as it leads to difficulties in measuring circularity and that the lack of a clear definition risks leading to the presentation of only win-win solutions (Corvellec et al., 2021). The latter makes it possible for firms to adopt CE with minimal changes (Kirchherr et al., 2017) or even greenwash their developments and thereby stabilize existing unsustainable economic organizations (Niskanen et al., 2020). This ambiguity also means that there is less coherence in the underlying assumptions compared to GPNs, as a theoretically mature research field strongly anchored in disciplinary research. One assumption of the CE literature is that the focus on material flows, the waste hierarchy, and growth opportunities, will be central in the future (Korhonen et al., 2018). Others focus more on the CE as a means to sustainable development (Geissdoerfer et al., 2017; Leipold et al., 2022), implying a broader goal for which the CE can be an instrument.
Inherent in the idea of the CE is that it develops at the micro- and meso-level in the form of local and regional economies. Yet, the macro level is increasingly gaining attention as the CE requires systemic changes (Kirchherr et al., 2023). In the development of a CE, actors’ intentions may misalign. For example, the promotion of a local- or regional CE may be constrained by policymakers operating at higher scales or by firms’ prioritizing a commercial view over spatial embeddedness (Newsholme et al., 2022), which would be instrumental for the development of regional CEs. Contributions from transition studies have underscored the often but not exclusively technology-led, changes in socio-technical systems, leading to more circular production and consumption patterns (Zepa et al., 2024) and further highlighted the role of local/regional policymaking in fostering local circular economies (Hartley et al., 2023; Haswell et al., 2024).
The literature on the CE brings together different disciplines such as industrial ecology, economics, and different social sciences, which further complicates the search for shared assumptions. According to Corvellec et al. (2021: 422), the CE concept suffers from “a perceived lack of paradigmatic strength” and research should therefore focus more on its core, that is, closed material loops and a fairer distribution of resources. Further limitations of the CE literature include: i) the neglect of established knowledge (Corvellec et al., 2021), ii) that the concept (as an instrument) will not deliver enough (e.g., related to biodiversity conservation) (Calisto Friant et al., 2020), iii) that the CE ignores material limitations of complete recycling (linked to thermodynamics laws) (Korhonen et al., 2018), iv) risks of shifting problems along the value chain or product life cycle and with different time horizons (i.e., rebound effects, Jevons paradox) (Korhonen et al., 2018), and v) its foundational conflicting ideological values (such as degrowth vs. capitalism) (Calisto Friant et al., 2020).
To summarize, the flora of scholarly definitions and diverse contributions to the CE scholarship opens for criticism from various directions. It suggests that the theoretical foundations of a CE need to be further developed (Calisto Friant et al., 2020), especially with regard to its theory of change (i.e., where and how change occurs). The idea that the CE is a means to sustainable development is shared among many, but the link between the two is criticized for being unclear and vague. When considering the concepts of CE and sustainable development as a complementary pair, the CE literature is criticized for prioritizing economic goals (Skene, 2017). Although the CE may bring environmental benefits, Calisto Friant et al. (2020) see risks of underestimating the importance of the social dimension (especially issues of governance, justice, and cultural change), when economic development dominate focus.
In search of a conceptual vocabulary to study a globally integrated circular economy
The analysis of assumptions underlying the two literatures demonstrates both similarities and differences. In terms of similarities, both literatures regard economic growth as foundational for development. The CE is perceived to provide a pathway for green growth by decoupling environmentally detrimental resource use from economic growth. Whether this is actually possible is contested, particularly in degrowth or post-growth approaches to the CE (Bauwens, 2021; Hickel and Kallis, 2019). Moreover, both literatures identify similar key actors: firms and states. Regarding firms, the GPN literature focuses mainly on MNCs while the CE literature also embraces actors like start-ups and intermediaries. The state is emphasized in the GPN literature as the one who sets the regulatory environment in which MNCs operate. Yet, it is also suggested in the literature that states and MNCs negotiate terms, foregrounding the importance of power relations between these actors. The CE literature instead emphasizes the state’s role as essential for providing incentives and regulations to promote the shift to circular business models. Finally, both literatures fall short of taking the role of the consumer seriously. In the case of GPNs, the analysis ends with the final product. Who uses the product, for how long, and what happens with the product when it is turned into waste is beyond the scope of the current GPN literature. However, the global trade of waste and its handling has been addressed in the adjacent and complementary literature on global recycling networks (e.g., Crang et al., 2012; Gregson and Crang, 2015). The CE literature embraces the whole life cycle of the product, from sourcing, production, and consumption to circulation. Yet, many contributions still underplay the role of the consumer and of historically institutionalized consumption practices through the focus on entrepreneurial and business activities as well as often considering consumers as passive adopters, rather than active agents (Hobson, 2015).
Key differences between CE and GPN research are first, GPN literature focuses on economic relations and value creation, whereas CE is predominantly concerned with material flows. By neglecting material flows, the GPN literature lacks the analytical toolbox to study material use and its associated environmental outcomes. Second, the temporal dimension: the work on CE is essentially prospective and forward-looking as there are currently few cases of a CE, notwithstanding rather small-scale and delimited experiments. GPN literature takes a retrospective perspective and investigates what has happened with the aim of identifying the causal mechanisms and drivers for the global integration of economic activities. Third, a central difference between the two literatures is the spatial understanding of production. The GPN literature takes a global approach, embracing a multi-scalar architecture of networks and institutions. Firms benefit from economies of scale, specialization, and a global division of labor. In a linear mindset, products are produced in locations with the most favorable conditions (e.g., in terms of cost or capabilities) and then shipped off to consumers. In a circular mindset, the scales of production, distribution, consumption, and circulation are often assumed to be smaller and regionalized. This mismatch in geographic focus between the two remains to be conceptually addressed, leading to questions such as: At what scales is it possible to create circular supply chains? How can the CE better incorporate multi-scalarity? Can circularity be realized within GPNs? Next, we scrutinize insights into these aspects by conceptualizing a globally integrated CE.
Conceptualizing a globally integrated circular economy
In this paper, we introduce the notion of a globally integrated CE, which we define as an economy that has established functions necessary for eliminating, reducing, or minimizing the use of (virgin) resources through resource loops organized at different spatial scales. We argue that it is necessary to theorize, conceptualize, and analyze the CE as globally integrated, despite the often implicitly held assumption that the local and regional scales are particularly important for economic relations in the CE. However, the argument for a globally integrated CE does not neglect the importance of local and regional resource loops or regional supply chains, but appreciates the global integration of the economy, which is the result of globalization embedded in a neoliberal regime since the 70s (Naidu et al., 2020). Globalization describes an increasing spatial division of labor (Massey, 1995) where not only nations but also regions develop narrow specializations and fulfill specific functions in GPNs (Dicken, 2015). The competitive dynamics inherent to the neoliberal regime are consequently solved by means of spatial fixes and a geographical expansion of economic activities (Harvey, 2001), with variegated social, economic, and environmental outcomes. Additionally, the policy and regulatory framework, which underpinned the era of neoliberal globalization, has been ignorant about the sustainable use of resources. As the extraction of resources, transport, energy, and the disposal of waste has been cheap and, in many places unregulated, globalization provided perfect incentives for the linear economy, consequently leading to negative environmental and social impact (e.g., Gregson and Crang, 2015; Meng et al., 2018; Ponte et al., 2023).
Functions of the globally integrated CE (based on and adapted from Reike et al., 2018; Coe and Yeung, 2015). These comprise of activities to meet a particular purpose in GPNs, respectively, GCNs and come with distinct resource flows.
The transition from a globally integrated linear economy to a globally integrated CE can be understood through the dynamic mechanism of what we call circular couplings. This describes the novel connections set up between regional economies providing new functions for the creation of closed resource loops and GPNs, respectively, decouplings from existing functions unnecessary for these loops (see next subsection). The resulting construct that emerges from these couplings is what we call a Global Circular Network (GCN). We use this term to account for the qualitative difference between GPNs and GCNs and the lack of existing conceptual vocabulary that considers relevant aspects such as actors, functions, and resources (see subsection on the geography of circular (de-)couplings) in the formation of these networks. While the resulting framework can be used to study industrial sectors such as automotive and textiles, we consider the framework broad enough to explain the functional dynamics of circular changes also in other sectors, given that some sort of material use is involved. We provide a conceptual vocabulary to study these changes that can inform empirical questions related to how, when, and why circular couplings differ across sectors and regions.
Circular (de-) couplings and functions
As we have outlined above, a globally integrated CE presents novel spatio-temporal configurations of economic processes that go beyond the realm currently discussed by the CE and GPN literatures. A central aspect of this shift is the mechanism of circular (de)coupling. Thereby, we start from the observation that a globally integrated CE will be formed through mechanisms similar to those that are described in GPN studies (see section on assumptions charaterizing the GPN literature). However, since the CE is a directed economic process, we consider circular couplings to comprise of processes by which the strategic direction of lead firms to pursue a CE (potentially driven by regulation) is aligned or matched with regional interests, resources, needs, and capabilities, including the circular missions of these regions and the activities of other actors than firms.
Building on the notion of couplings in GPN work, we define circular couplings as the dynamics of (dis)integration between local firms in regional economies which provide functions essential for closing resource loops, and MNCs operating within GPNs (see Figure 1). As outlined in the previous section, the GPN literature identifies different forms of couplings (functional, structural, and organic), based on the degree to which local firms are integrated into GPNs and whether these linkages originate from regional or extra-regional initiatives. The formation and leverage of couplings involve multiple actors, including policy actors operating at different spatial scales, as well as intermediaries and consultancies. Schematic depiction of the change from GPNs to GCNs involving the dynamic mechanism of circular (de)couplings that integrates regions 1 and 3 into a GCN, while decoupling region 2. Beyond the scales illustrated in this simplified figure, further coupling-relevant scales exist and are intentionally not displayed.
Against this backdrop, circular couplings can comprise of functional couplings, where regional functions (i.e., based on a certain set of techno-economic capabilities) match the economic interests and circular ambition of lead firms. Thereby, lead firms strategically couple with a region because of the availability of, for example, specific recycling or reuse facilities. 4 We can also think of endogenously driven organic circular couplings, where regional actors reach outside their context to create novel GCNs. 5 Circular couplings may also be a consequence of existing (structural) couplings being recoupled for other purposes, such as following CE regulation. This can relate to situations in which existing economic facilities of MNCs are getting repurposed for the CE. 6 Circular decouplings, on the other hand, describe situations in which regions and/or lead firms no longer consider existing couplings to be viable for the strategic goal of a CE, or consider other functions to be more relevant, and consequently cut them. This can relate to situations in which firms implement circular processes through which they substitute existing input materials. 7
Thus, taken together, these (de-)couplings represent necessary dynamics for the transition from a globally integrated linear economy to a globally integrated CE. Figure 1 illustrates a starting point where regions 1 and 2 contribute with three functions (A, B, C) to an existing GPN, which constitute the existing couplings (region 3 is not coupled). The transition to a GCN could entail that regions 1 and 3 succeed in making available and connecting other, potentially new functions, which are relevant for organizing circular resource flows. Such circular (re-) couplings (D, E) could entail things like resourcing, remanufacturing, or reuse. The transition from a GPN to a GCN may also entail that previous functions lose relevance and are thus disconnected, which would constitute a circular decoupling (C) in region 2.
But where and how can these circular couplings be observed? We argue that circular couplings are visible in processes by which new functions, necessary for the creation of closed resource loops against the background of a CE, are added to and integrated into existing GPNs or in the formation of emergent and novel GCNs. By functions, we mean techno-economic capabilities, encompassing what is discussed in the CE literature as different R’s (Reike et al., 2018). These R’s represent broad strategies (such as reuse or recycle) for creating resource loops. These strategies require that certain functions are put in place such as waste collection, waste trading, repair, rental or resell infrastructures. This can relate to a set of different or complementary functions being added for the circulation of resources (e.g., collecting and sorting infrastructure necessary for recycling), the reuse of products (rental companies providing products as a service), or the same waste stream being split up after the waste hierarchy (with parts being reused while others are recycled), correspondingly involving multiple couplings. According to Reike et al. (2018), the R’s can also be differentiated along resource loops. Short loops refer to situations in which the “product remains close to its user and function,” for example, through reuse. Medium loops describe instances in which “products are upgraded and producers are again involved,” such as in the repair or refurbishment of goods. Large loops refer to cases where “products lose their original function” (Reike et al., 2018: 254-255), for example, through the recovery of materials. Table 1 outlines the basic functions of GPNs and the CE to be considered in the formation of GCNs, the circular (de-)couplings involved, and their resource flows.
The geography of circular (de-)couplings
In this section, we synthesize the conceptual discussion so far by elaborating on the nexus of actors, functions, and resources in which circular resource flows are created and developed through circular couplings. We use the term nexus to signal that this involves an expansion of the state-firm nexus that is traditionally the focus of GPN literature and in which bargaining between state and firm rationalities takes place. Informed by existing literature, our model has three dimensions (Figure 2). The model can thus be used to empirically uncover the geography of circular (de-)couplings, that is, the places, spaces, and scales which are both conducive to and at which we experience effects of these couplings, involving specific actors, functions, and resources. Regions play a key role in these dynamics in valuing resources that flow into the region (through existing couplings), the processes through which they are transformed, upgraded/adapted, or (re)used in production and consumption in the region, and facilitating resource flows to production- and consumption-activities in other regions. While one might expect small resource loops (e.g., reuse) to align with local or regional scales, and large loops (e.g., material recovery) to have potential for global integration, our illustrations (see following subsection) will demonstrate that the geographical patterns of these couplings are far more heterogeneous. This complexity highlights the need for a sustained, long-term empirical research agenda focused on the geography of circular couplings. We will now first introduce the actors-functions-resources nexus, before returning to this issue more substantially in the discussion. The geography of circular (de)couplings results from the establishment and undoing of couplings in the CE. These couplings are analyzed through the actor-resources-functions nexus, which gives insights into the places and spatial scales conducive to and at which we can find effects of these couplings.
First, we incorporate a broader set of functions which include material transformation, upgrade or adaptation, and product (re)use, in addition to the canonical functions typically emphasized when studying global production (see Table 1). Economic geography research has a long history of studying functional dynamics in local and multi-scalar systems of innovation and production (see Asheim et al., 2016; Binz et al., 2016). Current frameworks typically distinguish between processes that lead to the build-up of new structures for resource provision, the realignment of existing structures, and the dismantling of structures that are no longer necessary to provide key industries with resources (Miörner and Trippl, 2019). We expect a shift towards more circular production patterns to lead to substantial reconfigurations of regional functional dynamics, as regional actors seek to exploit the opportunities and respond to the pressures that emerge when directionality changes in GPNs or green-field GCNs. It has also been demonstrated that context conditions (geographical- or field-/sector-related) affect how different actors value resources and visions about their future use (Jeannerat, 2024; Miörner, 2022).
Second, in terms of actors, the model includes a broad set of non-firm actors, such as consumers, as well as expands the role of firms and states in facilitating and coordinating resource flows, following insights in the CE literature. Actors’ strategies are shaped by geographical- and sectoral context conditions, and in turn shape circular (de)couplings in non-negligible ways. For example, regional actors may respond to external impetus created by actors elsewhere, by utilizing existing regional production- and innovation capabilities (structural couplings). Circular (de)couplings may also be driven inside-out by actors who push a circular agenda strategically in existing GPNs or novel GCNs (organic couplings) or actively engage in regional reconfiguration processes to substantially alter or create functions for the CE (functional couplings). Finally, the role of consumers as active parties in value chains demands firms to establish new relationships based on new business models, such as rental or product-service system designs (functional couplings). Following the literature on strategic agency in economic geography, we expect the types of agency to differ depending on the type of actor and the structural opportunities or challenges they are faced with (Grillitsch and Sotarauta, 2020).
Third, in terms of resources and resource flows, the scope is furthermore expanded from a focus on immaterial flows in GPN studies (i.e., knowledge, value) to include resource flows such as residual flows, spare parts, and (raw) material inputs. Table 1 indicates how circular (de)couplings relate to different types of resource loops, ranging from large (the function of products being repurposed and energy being extracted), and medium, to small loops (involving direct contacts with costumers and maintaining (some of) the functions of the original item) (Reike et al., 2018). The relational dimension of resource loops refers to the degree of functional integration between actors in different parts of the value chain. Through circular (de)couplings, we may see an increasing integration in terms of how suppliers of materials adapt their value-creating activities differently to consumers (e.g., by functional upgrading in a circular sense, moving along the spectrum of R’s), but this is not a necessary condition for a shift towards a more circular global economy. It can also be the case that suppliers reorient their activities and produce circular resources that can directly substitute linear ones in existing large resource loops.
Circular (de-)couplings: Operationalization and illustration
In the following, we illustrate the applicability of the actors-functions-resources nexus to uncover the geography of circular (de-)couplings with three examples, one each referring to material transformation, material adaptation, and (material) reuse (Table 1). These examples come from different sectors and illustrate different resource loops (large, medium, small), functional and actor dynamics as well as parts of the diversity of coupling types, creating distinct and heterogeneous geographical outcomes. The examples are summarized in Table 2 according to the conceptual categories of the nexus. It is possible to imagine many more potential combinations, and the conceptual and empirical relevance of the ones outlined here are subject to future studies, representing a long-term research agenda. Nevertheless, these illustrations show the applicability of our model to analyze the emerging phenomena of a globally integrated CE. (1) A case for material transformation in the automotive sector: Design and recycling by lead firm Summary of three distinct illustrative examples circular couplings, described through the actors -resources-functions-nexus, involving multiple spatial scales and resulting in distinct spatial dynamics.
(De)couplings related to material transformation can start from the notion that some resources, crucial to existing GPNs, are becoming increasingly scarce. This gives firms economic incentives to be attentive to what happens to their product after consumption. Combined with new regulations that require producers to take responsibility for discarded products, it pushes firms to engage with the question of remining, recovering, recycling and/or repurposing materials used in their production processes. For example, the End-of-Life Vehicle Directive in the European Union requires that new vehicles be recyclable or reusable up to 95% of their weight. As a result, multiple dynamics are currently observable. First, most European regions have developed new functions, authorized treatment facilities that extract resources from used vehicles, supporting lead firms to fulfill the directive (functional coupling). Used vehicles are dismantled, and materials are shredded and sorted for basic recycling. The technological capabilities required to recycle the material are low, meaning that in many regions the necessary regional functions for material remining and recycling can be built up. Due to the legal status of the directive, the valuation of the resulting materials is not dependent on local markets and exploited by entrepreneurs located in each national vehicle market. This type of coupling is characterized by large resource loops encompassing actor dynamics in which the authorized treatment facilities and the end-user are geographically disconnected (similar spatial dynamics as original GPN), and the products are repurposed. Shredded metals are sold on secondary markets for recycled metals and may end up in a range of industries other than vehicle manufacturing. As such, this type of circular coupling may link different manufacturing industries through overlapping resource loops.
Complying with increasing demands to recycle may also lead to changes in the design or production process, aimed to later simplify recycling and enabling large resource loops. Within the automotive industry this additional dynamic can be illustrated by Mega-Casting (a.k.a. Giga-Casting) where larger and more complex components are molded (instead of smaller components in potentially different materials which are welded together). Fewer components and a lower mix of materials enable the recycling process and increase the value of recycled materials. These changes to existing GPNs involve innovation and changes in production process of local firms (new functions such as die casting machinery) and include various functional (de)couplings. These may alter the number of actors in the GPN as some existing suppliers may become redundant, whereas others might strengthen their position (and upgrade). We expect the spatial dynamics to reflect the ones in the existing production networks, but with fewer lower tiers, resulting in regions becoming decoupled as Mega-Casting simplifies production process (Burggräf et al., 2024). (2) A case for material adaptation in the pram industry: Product service systems with new local actors
Circular couplings involving material adaptation can refer to initiatives and concepts that extend use of products, such as rental models or Product Service Systems (PSS) (Singhal et al., 2020). By developing new functions covering the maintenance, refurbishment, or repair of products, PSS reduces the need for new resources and promotes sustainable small and medium resource loops by making better use of existing products via durable and repairable design. Subscription and rental alternatives in consumer goods with temporal usage (children’s equipment, tools, etc.), for example, the rental of children products (prams, strollers, baby seats, and toys) is a growing trend in Europe (Koide et al., 2022: for an overview). With regards to actor dynamics, typically, new rental firms emerge in a local or regional context and consumers rent equipment from them as their baby grows. Cleaning and minor refurbishment as well as quality and safety checks allow for the extension of use for several user cycles. While most newly manufactured consumer products are produced in Asia, the services related to refurbishment and quality/safety checks need to be performed geographically closer to the consumer, requiring the functional coupling with local firms offering these functions. The geographical proximity is important to minimize transportation and last-mile logistics, since the impact of both transports and refurbishment otherwise diminishes the benefits from extended usage of the products (Nilsson and Hazen, 2025). With regards to spatial dynamics, this example creates an additional regional circulation of resources in PSS (geographically close to the consumption) that complements the existing both linear global production and consumption within GPNs. (3) A case for re(use) in the textile sector: Secondhand markets and platforms
Circular couplings involving reuse can refer to secondhand markets as in the case of textiles and fashion. Within the EU, regulatory developments, including the Textile Waste Directive implemented in 2025 and forthcoming extended producer responsibility (EPR) policy, are generating new imperatives for circularity in textiles. As of today, various platforms and markets facilitate the resale and reuse of secondhand fashion and textiles aiming to circulate resources. While traditional venues such as flea markets and (charity) secondhand shops are older and well-established models, the past decade has seen the rise of for-profit business models, with actors such as Sellpy, Vinted, and Depop, that are reshaping the secondhand textile market (Persson and Hinton, 2023). As part of the broader platform economy, these firms provide functions related to digital infrastructure and services that connect consumers and enable garment reuse at scale.
Initially, charity-based secondhand initiatives established organic couplings by setting up localized collection and sorting infrastructures (functions) and selling used garments through local retail outlets. Today, digital resale platforms introduce significant competition, operating primarily for profit (Persson and Hinton, 2023). The provision of digital services and trading infrastructure leverages functional couplings that attract interest from lead firms pursuing circular business models. With regards to actor dynamics, lead firms are engaging in these types of business as part of their circularity strategy. An example of this is H&M investing in the platform Sellpy (became majority shareholder). It benefits both parties as the platform gains access to H&M’s supply chain capabilities, while H&M benefits from Sellpy’s expertise in resale and reuse (H&M, 2021). By emphasizing reuse and resale, secondhand fashion models (stores and platforms) help establish small to medium-sized resource loops (products remain close to the original function) and connect with consumers. The spatial dynamics of charity and platforms are somewhat different. While early charity-based secondhand systems had a primarily local or regional focus, the globalization of used textile flows has shifted this pattern. Particularly for lower-value textiles, this expansion involves exports to African or Asian countries (Crang et al., 2012; EEA, 2025). In contrast, digital platforms typically concentrate on higher-value items and target consumers in Europe or the United States, resulting in national or multi-national resale geographies.
Discussion
Based on the argument that we need to consider the existing global architecture of the economy as a point of departure for circular changes; we have conceptualized the globally integrated CE through the notion of GCNs. This adds a new, previously neglected, stratum to the growing geographical literature on the CE (e.g., Bourdin and Torre, 2024; Davies et al., 2024) and contributes with a concept of directional change to the classical geographical study of economic processes of globalization (e.g., Coe et al., 2004; Dicken, 2015; Gregson and Crang, 2015; Henderson et al., 2002; Yeung and Coe, 2014). We have outlined how changes from GPNs to GCNs can be conceptualized through the mechanism of circular (de)couplings. This describes the integration of firms into GCNs following couplings with functions necessary for resource loops. To understand these processes and their spatial implications, we have outlined the actors-functions-resources nexus, where economic relationships can be reconfigured and transformed from linear GPNs to GCNs. We will now discuss implications and future research avenues in human geography arising from this.
The geography of circular couplings and global circular networks
With our framework, we go beyond the scales traditionally used in CE research (local, regional). This allows us to empirically study the scalar orientation of couplings, and which scales are conducive to specific forms of coupling across sectors. Furthermore, our framework makes it possible to visualize the changes in GPNs towards circularity and how they involve a variety of actors, including lead firms, regional firms, policymakers, consumers, etc. We argue that the conceptual vocabulary provided in this paper allows us to map the actor(s) that realize couplings as well as study the geography of power relations and shifts in these subsequent circular developments. We can assume that the scalar orientation of couplings differs greatly, depending on material realities of industries (sources for raw materials, possibilities to resource, remanufacture, or reuse products or components). Our illustrations present some of the various possibilities of actors, resources, and functions involved in couplings as well as the resource loops and spatial dynamics underpinning these. While one might expect to see smaller, circular resource loops operationalized at local or regional spatial scales, the textile example paints a much more heterogeneous picture and highlights the need to empirically scrutinize the scales of circular couplings. This not only concerns the scalar orientation but also refers to the scales that are conducive to specific types of couplings. For example, in the PSS illustration, functions such as repair and refurbishment must be available regionally in order to generate resource loops.
GPN literature foregrounds an understanding of regionally endogenous or exogenous actors, their integration in GPNs, and how firms and states bargain in different forms of structural, organic, and functional couplings. Circular couplings can, in principle, encompass all three of these dynamics. However, it will be important to understand which type of coupling dominates which sector, what functional dynamics and resource loops this encompasses, and whether lead firms will functionally couple with local firms or vertically integrate circular activities. The examples discussed in this paper do not capture instances of structural couplings; yet, such cases do exist empirically (e.g., Renault’s Re:factory, see Ellen Macarthur Foundation, 2021) and will be important to study for refining the framework. Given that our framework can be used to examine the operationalization of different circular visions (Calisto Friant et al., 2020), including sufficiency-oriented or degrowth approaches to the CE, questions as to how community actors can effectuate circular couplings demand attention. Future research could also explore the spaces in which actors interact in order to establish circular couplings, and examine the effect of existing path dependencies and routinized consumer behavior on these developments (see more details in Grillitsch et al., 2025b). Although the focus of this paper is less on value capture dynamics, it will be important to apply our framework to analyze these dynamics in circular couplings. This includes examining how scalar orchestrations intersect with value capture and whether small-scale actors are able to upgrade and capture value within the globally integrated CE (e.g., Gregson and Crang, 2015; Rülke et al., 2025).
Building on the notion of resources (material and energy), which has so far been neglected in GPN work, our framework also allows us to outline new links that emerge in the operationalization of the CE between existing GPNs. These links, their spatial dynamics, and how they may encompass more than just economic value, need to be further explored. In particular, this may also be a first attempt to bridge what has been discussed as global destruction, waste reduction networks, or global recycling networks (Bryson et al., 2024; Crang et al., 2012; Gregson and Crang, 2015; Herod et al., 2013) with the work of GPNs in considering the whole life cycle of products.
Leveraging circular couplings on the regional scale
Our illustrations show the heterogeneity of conducive geographical factors and subsequential effects on couplings. This calls for empirical research to nurture these factors, especially in light of states’ “strategization” 8 of economic development in the geo-economy (Babic, 2025) and upcoming implementations of CE legislation on EU level. Both could affect the spatial (re)location of economic activities (and functions) within GPNs, which could result in regional circular opportunities (for more details, see Stihl et al., 2025).
However, questions remain regarding how functions intersect with existing regional dynamics and drivers of inequality. We expect circular functions to grow in regions and nations where circular policies are implemented, and which can leverage the techno-economic resources to challenge existing path dependencies within GPNs. Decouplings, on the other hand, can be expected in regions which provide functions (e.g., virgin material extraction) that will be substituted by the circulation of resources. Given the labor intensity of many CE activities (e.g., collection and sorting of waste), it will be important to understand which emergent circular functions are located in which region, how this perpetuates existing or alters spatial divisions of labor, and whether and how regions with labor-intensive CE functions can upgrade to higher-value CE functions. This is especially relevant to consider since research on the global trade of waste has demonstrated how recycling and recovery activities historically have grown in regions with cheap labor and weak environmental regulation (Crang et al., 2012; Gregson and Crang, 2015; Gregson et al., 2011). It will therefore be important to examine the global implications of CE initiatives in the Global North and how they reconfigure the (global) trade of waste and its economic outcomes across space.
We suggest that insights from regional and innovation studies about (regional) path dependencies, agency, knowledge, capabilities, capacities, etc., will be important to better understand actors’ and regions’ ability to strategically couple into emerging GCNs or upgrade within these (e.g., Baumgartinger-Seiringer et al., 2022; Grillitsch and Sotarauta, 2020; MacKinnon, 2012). In general, circular couplings can provide economic development opportunities for regions, and regional innovation systems could be directed towards delivering on such outcomes, for example, by means of circular missions that provide directionality and guide the interaction of actors on the regional scale towards the development of circular functions. However, a lack of capacities and capabilities, as well as scale and scope economies, may give rise to (high value) functions located in agglomerations that have stronger innovation systems or easier market access. This comes with the risk of further perpetuating existing patterns of uneven development. We argue for an empirical research agenda to better understand how regions can break with path dependency and escape these patterns, what role the location of specific functions plays, and which forms of regional integration in GCNs are favorable for regional economic development. This will also assist in crafting solid and empirically situated policy proposals incorporating the different ways in which regional policy actors can support local firms in driving circular changes at different scales and coupling into emerging GCNs.
Conclusion
This paper has conceptualized the globally integrated CE through the lens of GCNs, introducing the mechanism of circular (de-)couplings to capture how transitions from linear GPNs unfold. By integrating perspectives on actors, resources, functions, and geography, we have outlined how changes towards circularity are embedded in existing global economic architectures and shaped by multi-scalar dynamics. The framework developed here draws attention to the geographical heterogeneity of circular change and the diverse constellations of actors involved in its orchestration. It highlights that circularity may come with different spatial dynamics depending on which functions are introduced, substituted, or relocated, and at which scale these processes occur. This suggests the need for empirical work that examines how conditions for circular couplings vary across sectors and regions, and how these differences affect opportunities for value creation, capture, and upgrading within emerging GCNs.
Future empirical work needs to focus on the nexus of circular functions, actor constellations, and resources (flows) and the resulting (de-)coupling and spatial dynamics. Methodologically, such work could entail a mapping of the relevant actors in an existing GPN as well as the new actor constellation in the emerging GCN, the circular (R) strategies the actors are pursuing, which new functions are required to accomplish these strategies, and how this relates to the use and flow of resources. We foresee the necessity to conduct concrete and situated studies because existing characteristics of the globally integrated linear economy, as well as opportunities and challenges for circular transitions, will differ by sectoral and territorial contexts. Consequently, this also calls for comparative studies that could differentiate between the more general drivers and the context-dependent specificities of why, how, when, where, and by whom a transition towards GCNs can be made possible. Comparative research could also be used to identify variations of transition pathways towards GCNs and analyze why and how they differ for instance in respect to outcomes for regional development, or in relation to their alignment with growth or degrowth paradigms. Going beyond such intensive research designs, we see the possibility for extensive research designs, investigating for instance how policy interventions relate to the emergence of certain circular functions, where they are located, and how this affects resource flows. Open data on resource flows, the material use and environmental impact in GPNs would be an important basis to evaluate progress towards closed resource loops, which could be linked to changing actor landscapes and provision of circular functions, and the economic outcomes for regional development.
We emphasize the need for empirical research, for which we provided some inspiration above, because transitions from GPNs to GCNs are highly configurational. This means that the outcomes cannot be predicted in advance but depend on the interplay of sets of preconditions and agentic processes. Irrespective of the details of specific transition processes, we argue, however, that the mechanisms will include different forms of circular (de-)couplings involving new spatial configurations of functions, actors, and resources (flows). Hence, the proposed conceptualization provides the analytical leverage to empirically explore the drivers and conditionalities, and to develop theories of change of such transitions.
Footnotes
Acknowledgments
The authors gratefully acknowledge the constructive feedback provided by the CIRCLE community, the Innovation Studies Group at the Copernicus Institute for Sustainable Development (Utrecht University), the Cirus Group at EAWAG, and participants at the International Sustainability Transitions Conference 2024 in Oslo, where earlier versions of this manuscript were presented. We also sincerely thank the editor and two anonymous reviewers for their valuable comments. Finally, Jonathan Friedrich and Linda Stihl contributed equally to this paper and are co-first authors.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The following funding contributed to this work: Research Excellence Programme on Agenda 2030 from Lunds Universitet Agenda 2030 [2022-204]; Svenska Forskningsrådet Formas [grant number 2022-00585]; Jan Wallanders och Tom Hedelius Stiftelse samt Tore Browaldhs Stiftelse [grant number P25-0063].
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.