The Circularity Paradox in Sustainable Supply Chains: A System Dynamics Framework of Rebound Effects, Value Leakage, and System Constraints

Abstract

Circular economy practices increasingly shape supply chains through reuse, repair, remanufacturing, recycling, resource recovery, and reverse logistics. However, greater circular activity does not necessarily produce proportional environmental, economic, or social benefits. Additional transportation, sorting, processing, and coordination requirements can erode recovery gains, while efficiency improvements may stimulate rebound demand. This study develops a literature-based qualitative system dynamics framework to explain these interacting mechanisms as the circularity paradox. Peer-reviewed literature and selected secondary evidence from the United States and Bangladesh inform the proposed causal relationships under different institutional and operational conditions. The framework connects circular intervention intensity, reverse flow complexity, recovery efficiency, organizational capability, processing capacity, secondary market absorption, value leakage, rebound demand, and sustainability performance. Three reinforcing and four balancing feedback loops illustrate how circular interventions may generate value creation, capability learning, operational burdens, capacity constraints, market saturation, and rebound effects. Eleven conditional propositions identify potential indicators, boundary conditions, and delayed responses for future empirical testing. Sustainable circular performance is treated as a multidimensional environmental, economic, and social outcome rather than a direct consequence of recovery volume. The framework suggests that sustainability gains depend on aligning circular activity with organizational, processing, and market capabilities.

1. Introduction

Circular economy principles have become increasingly important in efforts to reduce resource depletion, waste generation, and environmental pressure associated with conventional linear production and consumption. Rather than following a take-make-use-dispose model, circular systems seek to retain products, components, materials, and their embedded value through reduction, reuse, repair, refurbishment, remanufacturing, recycling, and resource recovery [1,2,3]. These principles have important implications for supply chain management because materials that previously left the economic system as waste are increasingly expected to return through reverse flows and re-enter production or consumption. Circular supply chain management therefore connects sourcing, production, distribution, consumption, collection, recovery, and secondary markets within a broader system of material and value flows [4,5,6]. The environmental logic supporting this transition is compelling, but the relationship between circularity and sustainability is not necessarily straightforward. Increasing collection, recycling, remanufacturing, or reuse can reduce waste and virgin resource requirements, yet the same activities may require additional transportation, sorting, storage, inspection, processing, energy, information exchange, and coordination. Returned products are often geographically dispersed and uncertain in quantity, timing, condition, and quality, creating operational requirements that differ substantially from conventional forward supply chains [6,7]. Recovered materials can also lose economic or functional value through contamination, degradation, inefficient processing, downcycling, or weak secondary demand. Thus, increasing the quantity of material circulating within a supply chain does not by itself establish that the supply chain has become more sustainable.

A further challenge is created by rebound effects. Efficiency improvements and circular business models may reduce the effective cost of products or resources, extend their availability, or create additional economic value. These benefits can alter producer and consumer behavior in ways that increase consumption and material throughput, partially offsetting the environmental improvements initially achieved [8,9,10]. Rebound can also occur at the system level when expansion of one circular activity creates additional requirements elsewhere. Higher collection volumes, for example, may increase recovery opportunities but can simultaneously place pressure on transportation networks, sorting facilities, processing capacity, storage infrastructure, and secondary material markets. The resulting sustainability outcome depends on the interaction between the circular intervention and the capacity of the wider system to support it. This study refers to this tension as the circularity paradox: increasing circular intervention can strengthen resource recovery and value retention while simultaneously activating operational, market, organizational, and behavioral mechanisms that weaken part of the expected sustainability gain. The paradox does not suggest that circular practices are inherently unsustainable. Rather, it emphasizes that circular activity and sustainable circular performance are not equivalent. A supply chain can become more circular in terms of material recovery while experiencing diminishing sustainability returns if the supporting system does not develop at a comparable rate.

Existing studies provide substantial knowledge about circular supply chains, reverse logistics, circular business models, organizational capabilities, recycling, remanufacturing, and circular economy barriers [6,11,12]. Another stream of research has identified circular economy rebound effects and demonstrated that circular interventions can produce unintended behavioral and economic responses [8,9,10,13]. However, these mechanisms are frequently examined separately. Less attention has been given to how rebound demand, reverse flow complexity, organizational capability, recovery efficiency, stakeholder coordination, and secondary market capacity interact as parts of the same supply chain system and collectively determine whether additional circularity creates or leaks sustainable value. Such relationships are inherently dynamic. Greater recovery may increase retained material value and strengthen investment in circular capabilities, creating a reinforcing process. At the same time, higher recovery volumes can increase reverse logistics complexity and processing pressure, creating balancing forces that constrain further improvement. Expanding secondary material supply may strengthen circular production when sufficient demand exists but depress recovered material value when market absorption is limited. Efficiency improvements may reduce resource requirements per unit while stimulating additional demand. These interacting processes involve feedback, delays, capacity constraints, and unintended consequences, making system dynamics particularly appropriate for examining the problem [14,15].

Accordingly, this study develops a literature-based qualitative system dynamics framework explaining how circular intervention intensity influences sustainable supply chain performance through reinforcing and balancing feedback processes. Selected secondary evidence from the United States and Bangladesh is incorporated to illustrate how these relationships may operate under different levels of infrastructure, organizational capability, market development, and institutional support. The two countries are not treated as directly comparable empirical cases; rather, they provide contrasting contexts for examining the broader theoretical relationships. The study addresses three questions: Why does increasing circular supply chain activity not necessarily produce proportional sustainability improvement? How do reverse flow complexity, organizational capability, recovery efficiency, stakeholder coordination, secondary market capacity, and rebound demand influence sustainable value creation and value leakage? What feedback structures determine whether circular interventions generate sustained benefits, diminishing returns, or unintended consequences? By addressing these questions, the study develops a system-level explanation of the circularity paradox and shifts attention from the amount of circular activity to the net sustainable value it generates. This distinction provides the foundation for the literature review and the qualitative system dynamics framework developed in the following sections.

2. Literature Review and Theoretical Foundations

Circular supply chain management extends circular economy principles across sourcing, production, distribution, consumption, return, recovery, and reintegration. Rather than treating waste as the final stage of a product life cycle, circular systems seek to retain products, components, materials, and embedded value through reduction, reuse, repair, refurbishment, remanufacturing, recycling, and resource recovery [1,2,3,16]. Circular supply chains consequently connect forward and reverse flows and require coordination among manufacturers, suppliers, customers, logistics providers, collectors, recyclers, remanufacturers, and secondary material users [4,5,6,11,17,18,19]. The expected benefits are substantial. Reuse and repair can extend product life, remanufacturing can preserve components and embedded value, and recycling can reduce disposal and substitute secondary materials for virgin resources. Circular practices may also improve resource productivity and create new economic opportunities when recovery activities are supported by appropriate organizational and institutional conditions [6,12,20,21]. However, the literature also increasingly recognizes that material circulation is not equivalent to sustainability. Circular strategies differ in the amount of functional, economic, and material value they retain, and recovery activities themselves consume resources [16,20,22,23,24]. This distinction between material circulation and value retention is central to the present study. Table 1 summarizes the major literature streams that provide the theoretical foundation for the circularity paradox.

Table 1. Major literature streams supporting the circularity paradox.

Literature Stream	Main Contribution	Relevance to This Study	References
Circular economy	Closing and slowing resource loops; reduction in waste and virgin-resource dependence	Establishes the broader circularity logic	Ghisellini et al. [1]; Geissdoerfer et al. [2]; Kirchherr et al. [3]; Korhonen et al. [16]
Circular supply chains	Integration of forward and reverse material flows	Establishes supply-chain-level circularity	Genovese et al. [4]; Batista et al. [5]; Govindan and Hasanagic [11]; Farooque et al. [6]
Value retention	Preservation of product, component, material, and economic value	Separates circulation from actual value preservation	Linder et al. [22]; Saidani et al. [23]; Velenturf and Purnell [20]; Muller et al. [24]
Reverse logistics	Collection, return, inspection, sorting, processing, and redistribution	Explains operational burden of circular flows	Govindan and Soleimani [7]; Batista et al. [5]; Farooque et al. [6]
Circular rebound	Circular improvements may stimulate additional consumption or fail to displace primary production	Explains counteracting sustainability effects	Zink and Geyer [8]; Makov and Font Vivanco [9]; Castro et al. [10]; Metic and Pigosso [13]; Schultz et al. [25]; Lowe et al. [26]; Ferrante et al. [27]
Dynamic capabilities	Organizations must integrate and reconfigure resources as conditions change	Explains differences in firms’ ability to manage circular complexity	Teece et al. [28]; Eisenhardt and Martin [29]; de Angelis et al. [30]
Digital circularity	Traceability, visibility, information sharing, and data-enabled coordination	Supports management of complex reverse flows	Nascimento et al. [31]; Kristoffersen et al. [32]; Chauhan et al. [33]; Taddei et al. [34]; Jum’a et al. [35]
System dynamics	Feedback, delays, capacity constraints, and unintended consequences	Provides the modeling foundation	Forrester [14]; Sterman [15]; Shamsuddoha et al. [36]; Shamsuddoha et al. [37]

Note: The literature streams overlap; references establish theoretical foundations and do not represent independent tests of the entire framework.

Their integration is necessary to explain how circular interventions can simultaneously create value and generate operational, market, organizational, and behavioral pressures.

2.1. Circularity, Value Retention, and Reverse Flow Complexity

Circular supply chains introduce operational characteristics that are less prominent in conventional forward supply chains. Returned products are uncertain in timing, quantity, location, condition, and quality. They may require collection, consolidation, transportation, inspection, disassembly, sorting, cleaning, storage, repair, remanufacturing, recycling, and redistribution before value can be recovered [5,6,7,38,39]. Increasing circular intervention therefore creates both a recovery opportunity and an operational burden. The distinction becomes important as circular activity expands. Early recovery may capture relatively accessible and high-value resources, whereas additional recovery can involve geographically dispersed sources, smaller quantities, greater contamination, or lower-value materials. Transportation, labor, energy, sorting, and processing requirements can consequently rise while the marginal value retained declines. This creates the possibility of value leakage, where part of the environmental or economic benefit expected from circularity is lost through the activities necessary to maintain the circular flow. The literature similarly identifies contamination, excessive transportation, inefficient sorting, low recovery yields, downcycling, storage delays, and insufficient processing capacity as mechanisms through which physical recovery and sustainable value recovery can diverge.

2.2. Rebound Effects and the Limits of Circular Expansion

Rebound provides a second explanation for why circularity and sustainability may diverge. Circular interventions can reduce resource requirements or effective costs, but the resulting savings can alter consumption, production, and market behavior. Zink and Geyer [8] show that circular economy rebound can occur when secondary products fail to displace primary production or when lower prices stimulate additional consumption. Makov and Font Vivanco [9] similarly demonstrate that rebound can materially affect the environmental gains attributed to circular interventions. More recent studies conceptualize circular rebound as a broader systemic phenomenon involving interacting behavioral, market, technological, and governance mechanisms [10,13,25,26,27,40,41]. Direct rebound can occur when lower costs encourage greater use of the same product or service. Indirect rebound can arise when savings are spent elsewhere, while economy-wide effects can influence prices, production, investment, and overall resource demand. Circularity may therefore improve resource efficiency per unit without producing an equivalent reduction in total material throughput. The literature also identifies direct, indirect, and wider system responses as important reasons that environmental improvement per unit may not translate into lower total environmental burden. The present study extends this reasoning beyond consumption rebound. A circular intervention may generate value leakage through transportation, processing, inventory, coordination, and capacity pressures even when consumer demand remains unchanged. Rebound demand and operational value leakage are therefore distinct but interacting mechanisms. Together, they explain how circular activity can increase while net sustainability benefits grow more slowly, stabilize, or, under unfavorable conditions, decline. Table 2 summarizes the principal mechanisms through which expected circular economy benefits can be weakened.

Table 2. Mechanisms that can weaken expected circular economy benefits.

Mechanism	Initial Circular Benefit	Counteracting Effect	References
Product reuse	Avoided or delayed new production	Lower prices or increased accessibility may stimulate additional consumption	Zink and Geyer [8]; Makov and Font Vivanco [9]
Remanufacturing	Retention of product and component value	Return uncertainty, inspection, disassembly, and processing requirements	Govindan and Soleimani [7]; Batista et al. [5]
Recycling	Virgin-material substitution and waste diversion	Energy use, sorting losses, contamination, and downcycling	Korhonen et al. [16]; Velenturf and Purnell [20]
Reverse logistics expansion	Greater resource capture	Additional transportation, handling, storage, and coordination	Govindan and Soleimani [7]; Farooque et al. [6]; Mishra et al. [38]; Mallick et al. [39]
Resource efficiency	Lower resource use per unit	Direct and indirect rebound may increase total throughput	Zink and Geyer [8]; Makov and Font Vivanco [9]; Castro et al. [10]
Increased recovery	Greater secondary resource availability	Processing capacity may become constrained	Shamsuddoha et al. [36]; Shamsuddoha et al. [37]
Secondary material growth	Greater potential virgin-material displacement	Supply can exceed market absorption	Zink and Geyer [8]; Velenturf and Purnell [20]
Circular scaling	Greater potential sustainability impact	Complexity can increase faster than organizational capability	de Angelis et al. [30]; Farooque et al. [6]; Metic and Pigosso [13]

Note: The counteracting effects are conditional mechanisms and should not be read as quantitatively estimated benefit reductions.

Table 2 demonstrates that counteracting effects can arise across multiple circular strategies, including reuse, remanufacturing, recycling, reverse logistics, resource efficiency, and secondary material recovery. These mechanisms provide the basis for incorporating value leakage and rebound effects into the proposed framework.

2.3. Capability, Coordination, and Market Absorption

Circular supply chains require capabilities for managing uncertain returns, material quality, multiple recovery pathways, information sharing, and relationships among organizations [17,18,42]. Dynamic capability theory is therefore relevant because firms must integrate and reconfigure resources as circular operations evolve [28,29,30]. Greater circular experience can generate learning and improve recovery efficiency, but circular activity can also expand faster than organizational capability. In that situation, delays, inventory accumulation, quality problems, and coordination costs increase. The literature supporting the present study specifically identifies this reciprocal relationship between circular activity, organizational learning, and operational complexity. Digital technologies can strengthen this capability through product identification, traceability, condition monitoring, data sharing, and improved visibility of reverse flows [31,32,33,34,35,43,44,45]. However, digital visibility does not remove physical constraints. Better information cannot independently create transportation capacity, sorting facilities, recycling technology, or secondary demand. Digital capability should therefore be viewed as one component of wider circular system capacity. Secondary markets represent another important constraint. Recovery creates sustainable value only when recovered products and materials can be reintegrated into economically viable applications. Material quality, consistency, technical standards, transaction costs, virgin-material prices, and buyer confidence influence this process. When recovered supply increases faster than demand, inventories can rise and recovered material prices can decline, weakening the economic case for further recovery investment [19,46,47]. Conversely, stable secondary demand can increase material value and encourage investment in quality and recovery capacity. Thus, recovery capacity and market absorptive capacity must develop together. Table 3 summarizes the principal system constraints and enabling conditions that influence sustainable circular performance.

Table 3. System constraints and enabling conditions for sustainable circular performance.

System Factor	Enabling Condition	Constraint When Misaligned	References
Organizational capability	Learning, adaptation, integration of circular processes	Complexity exceeds managerial capability	Teece et al. [28]; Eisenhardt and Martin [29]; de Angelis et al. [30]
Stakeholder coordination	Information sharing and aligned recovery decisions	Fragmentation and flow mismatch	Govindan and Hasanagic [11]; Farooque et al. [6]
Digital capability	Traceability, visibility, condition and flow information	Technology without physical capacity produces limited improvement	Nascimento et al. [31]; Kristoffersen et al. [32]; Chauhan et al. [33]; Taddei et al. [34]; Jum’a et al. [35]
Reverse logistics capacity	Efficient collection, consolidation, and transportation	Higher cost, delays, and dispersed flows	Govindan and Soleimani [7]; Farooque et al. [6]; Mishra et al. [38]; Mallick et al. [39]
Processing capacity	Sufficient sorting, remanufacturing, recycling, and treatment	Congestion, inventory accumulation, and recovery losses	Genovese et al. [4]; Shamsuddoha et al. [36]; Shamsuddoha et al. [37]
Recovered material quality	Consistent secondary resources suitable for productive use	Contamination and downcycling reduce retained value	Korhonen et al. [16]; Velenturf and Purnell [20]
Secondary market capacity	Reliable demand for recovered products and materials	Oversupply and declining recovery profitability	Zink and Geyer [8]; Velenturf and Purnell [20]
Demand response	Circular products substitute for primary production	Rebound increases consumption and total throughput	Makov and Font Vivanco [9]; Castro et al. [10]; Metic and Pigosso [13]

Note: Enabling and constraining conditions are theoretical categories that require sector-specific measurement before comparison.

Social sustainability is also part of circular system performance. Changes in collection and recovery can influence occupational safety, decent work, access to formal markets, the distribution of environmental burdens, and community participation. These effects should not be inferred from material recovery rates. The U.S. National Recycling Strategy explicitly considers worker safety, environmental justice, and equitable community access [48], while the World Bank identifies green skills and employment as potential dimensions of Bangladesh’s circular plastics transition [49]. These publications document policy priorities rather than realized social improvements. The framework therefore treats inclusion and worker protection as contingent social outcomes requiring separate indicators.

The relationships summarized in Table 3 indicate that sustainable circular performance depends on alignment across organizational, technological, logistical, processing, market, and demand conditions rather than on circular intervention alone.

2.4. Research Gap and Theoretical Position

The literature provides strong individual explanations for circular supply chain benefits, reverse logistics complexity, rebound effects, organizational capability, and secondary market constraints. Although recent studies have begun to model circular economy rebound using system dynamics, less attention has been given to an integrated supply-chain explanation that connects rebound with reverse-flow complexity, organizational capability, recovery capacity, secondary-market absorption, and operational value leakage as interacting feedback mechanisms. Reverse logistics research explains the difficulty of managing returns; capability research explains organizational adaptation; rebound research explains counteracting behavioral and market responses; and circular economy research explains resource retention. Examined separately, however, these streams provide only a partial explanation of net circular performance. This study therefore conceptualizes the circularity paradox through five connected mechanisms: resource recovery, value leakage, capability alignment, secondary market absorption, and rebound demand. Recovery creates potential value; organizational capability determines how efficiently that value can be retained; reverse flow complexity creates operational leakage; secondary markets determine whether recovered resources can be productively absorbed; and rebound responses determine whether efficiency improvements translate into absolute reductions in resource pressure. These mechanisms are mutually dependent rather than independent.

The resulting theoretical position is that circular intervention intensity and sustainable circular performance should not be assumed to have a continuously proportional relationship. Their relationship depends on the balance between value-creating and value-leaking mechanisms. Greater recovery can strengthen investment and capability, while greater complexity can increase operating burdens. Higher recovered-material quality can stimulate secondary demand, while excessive secondary supply can reduce material value. Efficiency can reduce resource requirements per unit, while rebound can increase total throughput. These opposing relationships create the reinforcing and balancing feedback structures examined through the system dynamics methodology in the following sections. Accordingly, this study conceptualizes sustainable circular supply chain performance as the outcome of interacting value-creating and value-leaking feedback processes, integrating recovery, reverse-flow complexity, organizational capability, capacity constraints, secondary-market absorption, and rebound within a qualitative system dynamics framework.

The novelty claimed here is not the discovery of circular economy rebound or an assertion that circularity always lowers sustainability. Earlier work already establishes rebound effects [8,9,10,13] and the limits of circularity indicators [16,20,22,23]. The more specific contribution is an integrated set of conditional, traceable feedback pathways linking demand-mediated rebound to operational value leakage, delayed capability and capacity adjustment, secondary market absorption, and multidimensional sustainability performance within one supply chain explanation. The resulting circularity paradox is a synthesis and theory-building construct, not a new empirically established law. This distinction specifies the incremental conceptual contribution and the boundary of the novelty claim.

3. Research Methodology

3.1. Research Design and Evidence Selection

This study adopts a literature-based qualitative system dynamics approach to examine the circularity paradox in sustainable supply chains [50]. System dynamics is appropriate for problems characterized by interdependence, feedback, delays, capacity constraints, and unintended consequences [14,15]. It has also been applied to supply chain and resource recovery systems to explain how changes in one part of a system influence economic and environmental outcomes elsewhere. Shamsuddoha [51] applied system dynamics to integrate forward and reverse supply chain processes in sustainable manufacturing, while Shamsuddoha et al. [36] modeled reverse logistical loops for poultry waste recovery. Subsequent studies extended this approach to sustainable dairy supply chains and dairy waste treatment, demonstrating how feedback structures, resource recovery, operational capacity, and alternative scenarios can be integrated within sustainability analysis [37,52]. Unlike those studies, the present research does not develop a calibrated simulation from primary operational data. It uses published literature and secondary evidence to construct a qualitative causal model explaining how circular interventions can simultaneously create sustainable value and generate counteracting pressures. Peer-reviewed studies on circular supply chains, reverse logistics, rebound effects, resource recovery, organizational capabilities, secondary material markets, and system dynamics form the principal evidence base. Recent literature is prioritized, while established studies are retained where necessary to support theoretical concepts and system dynamics methodology.

Evidence selection and reporting protocol. The submitted reference set was developed through a purposive, theory-oriented synthesis rather than a prospective systematic review. For this revision, the existing 61-entry reference set was retrospectively classified by its relevance to circular supply chain operations, rebound and behavioral responses, material value and markets, dynamic capabilities, qualitative system dynamics, or national policy context. Sources were considered suitable for a proposed causal link when they supplied a relevant theoretical mechanism, empirical observation, or established system dynamics convention. Irrelevant general commentary and claims without source support were excluded from the causal traceability register. Foundational studies were retained where conceptually necessary; newer works were used where present. This reconstruction is not presented as an exhaustive database search or PRISMA review. The documents supplied for revision do not contain original database search strings, dates, numbers screened, or exclusion logs. These historical search details cannot be independently reconstructed, and the original literature search is therefore not fully reproducible. However, the revised methodology provides an auditable conceptual evidence-selection process through explicit inclusion criteria, causal coding rules, and the traceability register in Appendix A.

Evidence coding and direction rules. A candidate relationship was retained if its source described a mechanism affecting at least one defined model variable and the effect could be stated as a directional ceteris paribus hypothesis. A positive link means that an increase in the cause tends to increase the effect relative to its otherwise expected level; a negative link means that it tends to decrease it. Literature support was distinguished from an additional author-inferred closing link. Where evidence allows several possible directions, the relationship is explicitly conditional rather than treated as universal. The codebook in Appendix A identifies each complete loop, the polarity sequence, and the principal references used to motivate its mechanism.

3.2. System Boundary and Model Development

The system boundary begins with circular intervention intensity and extends through reverse logistics, recovery operations, organizational capability, stakeholder coordination, secondary material markets, demand responses, and sustainability outcomes. Variables were included when the literature showed a recurring theoretical or operational relationship with circular supply chain performance. The main variables include circular intervention intensity, reverse flow complexity, recovery capacity, recovery efficiency, organizational capability, stakeholder coordination, recovered material quality, secondary material supply and demand, value leakage, rebound demand, virgin resource displacement, and sustainable value creation. The causal loop diagram (CLD) was developed in three stages. First, recurring relationships among the selected variables were identified from the literature. Second, the direction of each relationship was specified. A positive relationship indicates that two variables move in the same direction, all else being equal, while a negative relationship indicates movement in opposite directions. Third, the relationships were connected into reinforcing and balancing feedback loops. Reinforcing loops amplify an initial change, whereas balancing loops constrain growth or counteract the original change [15]. The model development also builds on the systems logic used in earlier supply chain applications by Shamsuddoha and colleagues, where causal relationships were translated into feedback structures before evaluating alternative system behavior. Applications to poultry supply chains, dairy operations, and dairy waste recovery demonstrate how economic, environmental, logistical, and recovery variables can be represented within an interconnected system rather than analyzed independently [36,37,51,52]. The present study extends this methodological logic from sector-specific operational systems to a broader theoretical model of circular supply chain performance.

Model construction and boundary choices. The model boundary follows circular intervention decisions, material returns, processing, market absorption, investment and demand effects through three separate sustainability dimensions. Construct names were standardized so each causal arrow has one source, one target and one stated direction. Candidate chains were first written as ordered causal relationships, then checked for a return path to the starting variable. A reinforcing loop requires an even number of negative links; a balancing loop requires an odd number. Links that previously yielded an open chain or contradicted the intended interpretation were revised, or the loop was respecified as a conditional path. The resulting framework consists of seven closed feedback circuits that represent interacting reinforcing and balancing mechanisms within the circular supply chain system. The loops share variables and mechanisms and are not seven independent empirical subsystems.

3.3. Qualitative Analysis and Model Credibility

Because the study is conceptual and based on secondary evidence, the causal relationships represent theoretically expected directions rather than statistically estimated effects. The model identifies conditions under which reinforcing mechanisms strengthen circular performance and conditions under which balancing mechanisms create diminishing returns, value leakage, capacity pressure, or rebound effects. Selected evidence from the United States and Bangladesh is used to illustrate these mechanisms under different supply chain and institutional conditions rather than to conduct a direct country comparison. The structural credibility of the proposed qualitative model was assessed through consistency with supporting literature, causal-link traceability, feedback-loop closure, and polarity checks. These procedures assess internal coherence rather than empirical or behavioral validation. Relationships are retained when supported by established theory, peer-reviewed empirical findings, prior system dynamics research, or clear material-flow and supply chain logic. This approach is consistent with the use of system dynamics for representing complex sustainability relationships before quantitative calibration [15]. Future research can convert the qualitative structure into a stock-and-flow simulation, estimate parameters, test alternative policies, and conduct sensitivity and behavioral validation.

Structural credibility procedures performed during revision comprise a consistency check of the proposed causal relationships and feedback structures; an arithmetic polarity check of all seven closed paths; inspection for contradictory labels or reversed oversupply effects; and comparison of proposed directions with the cited literature and secondary context. These procedures assess logical coherence and source traceability, not empirical validity. The materials supplied for revision provide no documentation of an external expert panel, interview-based validation, primary material-flow verification, calibrated simulation, or historical behavioral test; none is claimed as completed here. Independent expert review and empirical validation of structure and behavior remain future research requirements.

3.4. Stock, Flow, Delay, and Nonlinear Structure

A qualitative causal loop diagram does not itself constitute a calibrated stock-and-flow model. The proposed structure nevertheless has explicit stock and flow interpretations. Candidate accumulations include unprocessed returned products, secondary material inventory, installed effective processing capacity, and accumulated organizational capability. Corresponding flows include return arrivals and processing completions, secondary material recovery and sales, capacity commissioning and retirement, and capability learning and depreciation. For example, unprocessed returns change by arrivals less completions; secondary inventory changes by recovered output less market uptake. These are proposed conservation relationships, not measured equations or simulated outputs. Their development into numerical level and rate equations requires future operational data. These stock-and-flow interpretations identify candidate state variables for future quantitative modeling; they do not imply that every variable in the proposed causal loop model is a stock or that the proposed accumulation relationships have been empirically parameterized.

Delays are expected between a perceived capacity shortfall and commissioning new capacity, between circular operating experience and capability accumulation, and between changes in effective product costs and realized rebound demand. Congestion may increase nonlinearly as utilization approaches effective processing capacity, while recovered material prices may respond more strongly when supply persistently exceeds secondary demand. These threshold-like patterns are theoretical and context-dependent; the present study estimates neither their values nor their functional forms. The proposed balancing capacity path explicitly incorporates delayed adjustment. Propositions P10 and P11 make the delay implications testable using future time-indexed measures.

4. Development of the Circularity Paradox Framework

The literature indicates that circular supply chain performance is shaped by the interaction of resource recovery, reverse logistics, organizational capability, processing capacity, market demand, and behavioral responses. Circular interventions can create value by extending product life, recovering materials, reducing waste, and displacing virgin resources. However, the same interventions can increase transportation, sorting, processing, storage, coordination, and energy requirements. The central argument of the proposed framework is therefore that greater circularity creates sustainable value only when the supporting system can absorb the additional operational and market requirements generated by circular flows [4,6,10].

4.1. Circular Intervention and Sustainable Value Creation

Circular intervention intensity represents the extent to which a supply chain applies reuse, repair, refurbishment, remanufacturing, recycling, resource recovery, and reverse logistics. At lower levels of circular activity, increasing intervention can capture products and materials that would otherwise be discarded. This can reduce waste, preserve embedded value, and decrease demand for virgin resources [2,20]. Greater recovery can also generate revenue and organizational learning, supporting further investment in circular infrastructure and capabilities. The sustainability benefit of additional recovery, however, depends on what happens after products and materials enter the reverse supply chain. Returned resources differ in location, quantity, condition, composition, and quality. Increasing recovery can therefore expand transportation distances, sorting requirements, inventory, processing intensity, and coordination needs [6,7]. As increasingly difficult or lower-value resources enter the recovery system, the marginal benefit of additional circular activity may decline. This study describes the resulting loss as value leakage. Value leakage occurs when part of the environmental or economic value expected from a circular intervention is lost through additional logistics, processing, energy consumption, contamination, degradation, downcycling, coordination costs, capacity limitations, or weak market demand. The concept connects circular supply chain operations with rebound thinking by showing that sustainability gains can be reduced even when physical recovery continues to increase. Table 4 defines the core variables used in the circularity paradox framework and identifies their roles within the system.

Table 4. Core variables and their roles in the circularity paradox.

Variable	Definition in the Framework	Role in the System	References
Circular intervention intensity	Extent of reuse, repair, remanufacturing, recycling, recovery, and reverse logistics	Initiates and expands circular flows	Geissdoerfer et al. [2]; Farooque et al. [6]
Reverse flow volume	Quantity of products, components, and materials returning to the supply chain	Creates recovery opportunity but increases handling requirements	Govindan and Soleimani [7]; Farooque et al. [6]; Mishra et al. [38]; Mallick et al. [39]
Reverse flow complexity	Uncertainty and heterogeneity in timing, location, condition, and quality of returns	Increases logistical and processing burden	Govindan and Soleimani [7]; Batista et al. [5]
Organizational capability	Ability to integrate, coordinate, learn, and adapt circular operations	Determines ability to manage increasing circular complexity	Teece et al. [28]; de Angelis et al. [30]
Stakeholder coordination	Alignment and information sharing among actors involved in circular flows	Improves visibility, collection, and recovery	Govindan and Hasanagic [11]; Farooque et al. [6]
Recovery capacity	Available collection, sorting, treatment, recycling, and remanufacturing capability	Determines how much reverse flow can be effectively processed	Genovese et al. [4]; Shamsuddoha et al. [36]
Recovery efficiency	Ability to convert returned resources into usable products, components, or materials	Increases retained value and reduces losses	Velenturf and Purnell [20]; Shamsuddoha et al. [36]
Secondary market absorptive capacity	Ability of markets to utilize recovered products and materials	Determines whether recovered resources create economic value	Zink and Geyer [8]; Castro et al. [10]
Rebound demand	Additional consumption or throughput associated with efficiency, availability, or cost effects	Can offset environmental gains	Zink and Geyer [8]; Makov and Font Vivanco [9]; Castro et al. [10]
Value leakage	Loss of expected circular value through operational, environmental, or market mechanisms	Reduces net benefit from circular activity	Korhonen et al. [16]; Velenturf and Purnell [20]
Sustainable circular performance	Multidimensional environmental, economic, and social performance profile, assessed against a defined baseline; not a computed composite score.	Primary outcome profile; net value is interpreted within, not automatically summed across, sustainability dimensions.	Korhonen et al. [16]; Farooque et al. [6]
Social performance	Worker safety, decent work, meaningful stakeholder participation, and access to circular opportunities.	Tracks distributional and workforce consequences alongside environmental and economic outcomes.	U.S. EPA [48]; World Bank [49]

Note: Variables and social outcomes are conceptual. Their indicators and measurement units require operational definition in future studies.

Construct differentiation. Operational value leakage refers to lost potential circular benefits resulting from the activities required to recover and circulate materials, such as transport emissions, additional handling, processing losses, unsafe work, and capacity-related deterioration. Circular rebound is different: it occurs when circularity or efficiency changes prices, access, or behavior and thereby induces additional demand or material throughput that offsets expected resource savings [8,9,10,13]. Value leakage can therefore occur with unchanged final demand, whereas demand-mediated rebound requires a consumption or production response. Neither term should be treated as synonymous with simple waste loss, financial expenditure alone, or a guaranteed negative effect. The mechanisms can interact because capacity pressure can weaken value recovery while low effective costs stimulate throughput.

Outcome assessment. Sustainable circular performance is treated as a three-dimensional outcome profile, not a single monetized index and not an unweighted addition of environmental burdens, monetary costs, and social outcomes. The environmental dimension concerns absolute material and energy use, greenhouse gas emissions, waste disposal, and verified virgin-resource displacement; the economic dimension concerns recovery yield, unit processing and logistics cost, recovered material value, inventory carrying costs, and investment viability; the social dimension concerns worker safety, job quality and inclusion, community access, and distribution of local burdens. The phrase net sustainable value refers to remaining benefits within each dimension after its relevant counteracting burdens, without assuming automatic convertibility across units. Future aggregate comparisons would require an explicit system boundary, functional unit, baseline, normalization, weights and stakeholder justification; these are not supplied or estimated here.

Collectively, the variables in Table 4 distinguish circular activity from its net sustainability outcome. Their interactions determine whether additional circular intervention strengthens sustainable value creation or increases value leakage and counteracting pressures.

4.2. Capability and Capacity Constraints

Circular intervention places additional demands on organizational capabilities, and recent studies show that organizational, technological, institutional, and market barriers often interact rather than operate independently [46,53,54,55]. Firms must coordinate product returns, evaluate material condition, determine appropriate recovery pathways, exchange information with external partners, and adapt operations to uncertain reverse flows. Dynamic capability theory suggests that performance depends on the ability of organizations to integrate and reconfigure resources as environmental and operational conditions change [28,29]. In circular supply chains, this means that expansion of circular activity needs to be accompanied by corresponding development of logistical, technological, informational, and managerial capabilities. When capability develops alongside circular activity, learning and experience can improve recovery efficiency and support further circular investment [56,57]. When circular intervention expands faster than capability, however, a capability gap emerges. Increasing reverse flows can then create congestion, inventory accumulation, processing delays, contamination, coordination problems, and higher operating costs. Similar capacity relationships have been observed in system dynamics applications of resource recovery, where expansion of recovery activities without sufficient processing and logistical capability can shift system bottlenecks rather than eliminate them [36]. Physical recovery capacity creates a related constraint. Increased collection produces sustainability benefits only when sufficient sorting, recycling, remanufacturing, treatment, and storage capacity is available. When reverse flows approach or exceed available capacity, delays and congestion can reduce recovery efficiency and recovered material quality. Because infrastructure expansion requires investment and time, a delay can develop between growth in circular intervention and the capacity required to support it. This creates an important balancing mechanism within the circular supply chain.

4.3. Secondary Markets and Rebound Pressure

Recovery alone does not complete a circular flow. Recovered products and materials must be absorbed by markets capable of returning them to productive use. When secondary demand is strong, increased recovery can improve revenue, encourage investment, and strengthen recovery capability. When recovered material supply grows faster than demand, however, prices can decline, inventories can accumulate, and the financial attractiveness of recovery can weaken. Secondary market absorptive capacity therefore establishes another boundary on the scale at which circular intervention can generate sustainable value [8,20]. Circular interventions can also alter demand. Reuse, remanufacturing, sharing, recycling, and efficiency improvements can reduce effective costs or increase product availability. Consumers or firms may respond by increasing consumption, while financial savings may be redirected toward other resource-consuming activities. These responses can reduce the amount of virgin production or environmental burden actually displaced by circular strategies [8,9,10]. Rebound pressure is therefore incorporated into the framework as a mechanism linking circular efficiency with total system throughput. Table 5 contrasts the potential value-creating effects of major circular mechanisms with the value leakage or counteracting effects that may accompany them.

Table 5. Value-creating and value-leaking mechanisms in circular supply chains.

Circular Mechanism	Potential Value Creation	Potential Value Leakage or Counteracting Effect	References
Reuse and repair	Extends product life and avoids new production	Collection, inspection, repair, and redistribution burdens	Geissdoerfer et al. [2]; Zink and Geyer [8]
Remanufacturing	Preserves components and embedded value	Disassembly costs, uncertain returns, and quality variation	Govindan and Soleimani [7]; Batista et al. [5]
Recycling	Reduces disposal and can displace virgin materials	Energy use, contamination, processing losses, and downcycling	Korhonen et al. [16]; Velenturf and Purnell [20]
Reverse logistics expansion	Captures additional products and materials	Transportation, storage, handling, and coordination requirements	Govindan and Soleimani [7]; Farooque et al. [6]; Mishra et al. [38]; Mallick et al. [39]
Increased recovery volume	Improves resource recovery and may create economies of scale	Capacity congestion and declining marginal recovery value	Genovese et al. [4]; Shamsuddoha et al. [36]
Organizational learning	Improves circular capability and recovery decisions	Benefits may lag behind rapid expansion of circular activity	Teece et al. [28]; de Angelis et al. [30]
Stakeholder coordination	Improves information, recovery quality, and flow alignment	Requires governance, information sharing, and coordination resources	Govindan and Hasanagic [11]; Farooque et al. [6]
Secondary material supply	Reduces dependence on virgin resources	Oversupply can weaken prices and recovery profitability	Zink and Geyer [8]; Velenturf and Purnell [20]
Resource efficiency	Reduces resource requirements per unit	Lower effective cost can stimulate rebound demand	Makov and Font Vivanco [9]; Castro et al. [10]

Note: Value-creating and value-leaking effects are potential mechanisms, not measured net impact or a transferable numerical ranking.

4.4. Integrated Causal Structure

The proposed framework integrates these mechanisms into a system in which circular intervention produces both value-creating and value-constraining pathways, consistent with recent systems-oriented circular economy research [50,58,59]. The first pathway operates through recovery, value retention, reduced virgin resource dependence, organizational learning, and circular investment. The second operates through reverse flow complexity, operating burden, capacity pressure, value leakage, secondary market saturation, and rebound demand. These interacting pathways are conceptualized in Figure 1.

Figure 1. Conceptual structure of the circularity paradox in sustainable supply chains. Source: Authors’ development based on Zink and Geyer [8], Govindan and Soleimani [7], Farooque et al. [6], Velenturf and Purnell [20], Castro et al. [10], and Shamsuddoha et al. [36]. Note: Figure 1 is a conceptual process overview. Arrows show hypothesized relationships rather than measured causal magnitudes or validated feedback-loop closures; individual closed feedback paths are examined in the subsequent analysis.

The conceptual structure suggests that the relationship between circular intervention and sustainability is conditional rather than uniformly positive. When organizational capability, processing capacity, stakeholder coordination, and secondary market demand develop alongside circular intervention, the value-creating pathway can dominate. When circular intervention expands more rapidly than these supporting conditions, value leakage and rebound mechanisms become stronger. Table 6 summarizes the expected sustainability consequences under different combinations of circular activity, supporting capability, processing capacity, market conditions, and rebound pressure.

Table 6. Expected system conditions and sustainability consequences.

System Condition	Circular Activity	Supporting Capability and Capacity	Expected Sustainability Consequence	References
Early circular development	Low to moderate	Capacity generally sufficient	Strong initial gains from additional recovery	Geissdoerfer et al. [2]; Farooque et al. [6]
Coordinated expansion	Moderate to high	Capability and markets develop with recovery	Sustained value creation and higher recovery efficiency	de Angelis et al. [30]; Velenturf and Purnell [20]
Capacity-constrained expansion	High	Processing or logistics capacity develops slowly	Congestion, higher cost, and increasing value leakage	Govindan and Soleimani [7]; Shamsuddoha et al. [36]
Market-constrained circularity	High	Secondary demand remains limited	Inventory accumulation and declining recovered material value	Zink and Geyer [8]; Velenturf and Purnell [20]
Rebound-dominated circularity	High efficiency/circularity	Demand response is insufficiently controlled	Higher throughput offsets part of expected environmental gain	Makov and Font Vivanco [9]; Castro et al. [10]

Note: These conditions are qualitative analytical scenarios, not calibrated simulations or predicted transitions.

These relationships establish the causal foundation for the feedback structure. The next stage connects them into reinforcing and balancing loops to explain how the circularity paradox develops over time.

5. Feedback Structure and Causal Loop Model

The circularity paradox becomes clearer when the relationships identified in Section 4 are represented as interacting feedback loops. The proposed causal loop model contains three reinforcing loops and four balancing loops. The reinforcing loops explain how recovery value and reinvestment, capability learning, and secondary market development can strengthen circular performance. Stakeholder coordination is treated as a broader enabling condition rather than a separate reinforcing feedback loop in Figure 2. The balancing loops capture reverse flow complexity, capacity pressure, market saturation, and rebound effects that can reduce the net sustainability gains from expanding circularity.

Figure 2. Audited closed feedback paths in the circularity paradox framework (R1–R3 and B1–B4). Note: Each arrow is marked (+) for a same-direction response or (−) for an opposite-direction response. The link sign refers to change relative to an otherwise expected level, not to a measured correlation. Every loop returns to its starting variable; an even number of negative arrows indicates reinforcing polarity and an odd number indicates balancing polarity. The seven displayed paths share recurring constructs and form an interdependent conceptual structure. Loop closure and causal directions are conditional hypotheses rather than empirical validation. Source: Authors’ development based on Sterman [15], Zink and Geyer [8], Govindan and Soleimani [7], Farooque et al. [6], Makov and Font Vivanco [9], Velenturf and Purnell [20], Shamsuddoha et al. [36], Shamsuddoha et al. [37], and Castro et al. [10].

5.1. Reinforcing Feedback Loops

R1: Recovery–Investment Loop. Greater circular intervention increases resource recovery, which increases the quantity of products, components, and materials available for productive reuse. When these resources retain sufficient quality and market value, recovery generates economic returns that strengthen the business case for further circular investment. Additional investment improves recovery capability and supports further circular intervention. This reinforcing process helps explain how economically viable recovery systems can become progressively stronger once sufficient scale and market value are established [4,6,20].

In its closed form, retained material value supports recovery profitability, which supports circular investment; investment strengthens recovery capability and thereby subsequent recovery. This return link is conditional on sufficient secondary demand and the reinvestment of circular returns.

R2: Capability–Efficiency Loop. Circular intervention also generates operational experience and organizational learning. Learning strengthens circular supply chain capability, which improves decisions concerning collection, sorting, processing, reuse, remanufacturing, and recycling. Higher capability increases recovery efficiency and retained value, strengthening incentives for continued circular investment. This relationship is consistent with dynamic capability theory, in which organizational performance depends on the ability to integrate and reconfigure resources as operating conditions change [28,29,30].

For closure, investment increases subsequent circular activity and experience, while experience develops organizational capability. The capability stock accumulates with learning and can deteriorate through staff turnover or obsolescence; the diagram depicts only the hypothesized net positive learning link.

R3: Quality–Market Development Loop. Improved sorting and recovery capability increases the consistency and quality of recovered products and materials [60,61]. Higher quality increases buyer confidence and secondary market demand, which improves recovered material value and strengthens investment in recovery technology. Improved technology subsequently increases recovered material quality. This loop demonstrates that secondary markets are not external to circular supply chains but can co-evolve with recovery capability and material quality [8,20].

The causal path closes when recovered material value sustains recovery investment, investment strengthens recovery capability, and capability supports quality improvements. The direction is conditional on buyer confidence, quality reliability, and actual market uptake.

5.2. Balancing Feedback Loops

B1: Reverse Flow Complexity Loop. Increasing circular intervention expands the quantity and diversity of reverse flows. Greater heterogeneity and geographic dispersion increase collection, transportation, inspection, sorting, inventory, and coordination requirements. These requirements increase circular operating costs and value leakage, reducing the net return from additional circular activity. The loop therefore constrains expansion when reverse flow complexity increases faster than organizational capability [6,7].

The balancing path returns from lower net circular return to reduced incentives for circular investment and therefore to lower subsequent intervention intensity; this negative feedback is distinct from the direct physical burden.

B2: Delayed Capacity Adjustment Loop. As unprocessed returns accumulate, processing congestion makes the effective capacity shortfall more visible. The resulting perceived gap may motivate capacity investment; after an implementation delay, commissioned processing capacity increases completions and reduces the outstanding backlog. This is a balancing capacity-adjustment path with one negative closing link, from effective capacity to unprocessed returns. During the investment delay, congestion, sorting losses, and inventory-related value leakage may worsen rather than improve immediately [15,36,37].

B3: Secondary Market Saturation Loop. Greater recovery increases the supply of secondary products and materials. When supply grows faster than secondary market demand, market absorption declines and recovered material prices can weaken. Lower recovered material value reduces recovery profitability and investment incentives, constraining further expansion. The effect is particularly important when recovered materials compete directly with relatively inexpensive virgin resources [8,20].

Here, oversupply is defined as recovered supply relative to demand, so the link from oversupply imbalance to recovered material price is negative. Lower prices reduce recovery profitability and investment; weaker investment eventually constrains secondary material supply. The closing loop is balancing, not a positive relationship between oversupply and value.

B4: Sustainability Rebound Loop. Improvements in resource efficiency, reuse, remanufacturing, or other circular practices can reduce effective product or resource costs. Lower costs and greater availability may increase consumption and total throughput. As throughput rises, production, transportation, energy consumption, and material requirements can also increase, offsetting part of the environmental benefit initially generated by circular intervention. The magnitude of this effect depends on consumer behavior, market conditions, substitution between secondary and primary products, and the extent to which circular products replace rather than supplement new production [8,9,10]. Table 7 consolidates the three reinforcing and four balancing feedback processes, their principal causal pathways, and their implications for circular supply chain performance.

Table 7. Feedback loops in the circularity paradox framework.

Loop	Type	Main Causal Pathway	System Implication	References
R1 Recovery–Investment	Reinforcing	Resource recovery (+) → Retained material value (+) → Recovery profitability (+) → Circular investment (+) → Recovery capability (+) → Resource recovery [closure]	Successful recovery can become self-reinforcing	Genovese et al. [4]; Farooque et al. [6]; Velenturf and Purnell [20]
R2 Capability–Efficiency	Reinforcing	Circular experience (+) → Organizational capability (+) → Recovery efficiency (+) → Retained material value (+) → Circular investment (+) → Circular experience [closure]	Learning strengthens the ability to manage circular flows	Teece et al. [28]; Eisenhardt and Martin [29]; de Angelis et al. [30]
R3 Quality–Market Development	Reinforcing	Recovery capability (+) → Recovered material quality (+) → Buyer confidence (+) → Secondary demand (+) → Recovered material value (+) → Recovery investment (+) → Recovery capability [closure]	Quality and market development can strengthen each other	Zink and Geyer [8]; Velenturf and Purnell [20]
B1 Reverse Flow Complexity	Balancing	Circular intervention intensity (+) → Reverse flow complexity (+) → Operating burden (+) → Operational value leakage (−) → Net circular return (+) → Circular investment (+) → Circular intervention intensity [closure]	Complexity can reduce marginal benefits of additional circularity	Govindan and Soleimani [7]; Farooque et al. [6]; Mishra et al. [38]; Mallick et al. [39]
B2 Delayed Capacity Adjustment	Balancing	Unprocessed returns stock (+) → Processing congestion (+) → Perceived capacity gap (+) → Capacity investment (+) → Effective processing capacity (−) → Unprocessed returns stock [closure]	Capacity response can reduce backlog only after an investment and commissioning delay.	Shamsuddoha et al. [36]; Shamsuddoha et al. [37]
B3 Secondary Market Saturation	Balancing	Secondary material supply (+) → Supply–demand oversupply (−) → Recovered material price (+) → Recovery profitability (+) → Recovery investment (+) → Secondary material supply [closure]	Oversupply decreases recovered material price, slowing subsequent supply expansion.	Zink and Geyer [8]; Velenturf and Purnell [20]
B4 Sustainability Rebound	Balancing	Circular efficiency (+) → Effective cost savings (+) → Rebound demand (+) → Total material throughput (+) → Environmental burden (−) → Net sustainability performance (+) → Circular investment (+) → Circular efficiency [closure]	If investment depends on net performance, rebound counters subsequent efficiency expansion.	Zink and Geyer [8]; Makov and Font Vivanco [9]; Castro et al. [10]

Note: The sign following each variable indicates the polarity of its outgoing causal link to the next variable. Positive signs indicate same-direction relationships, whereas negative signs indicate opposite-direction relationships, all else being equal. Loop classifications were checked against Figure 2 and Appendix A. The pathways are conceptual and have not been empirically validated.

This balancing pathway is closed conditionally through net sustainability performance and reinvestment: circular efficiency produces effective savings, savings can induce rebound demand, increased throughput increases environmental burden, and that burden reduces measured net sustainability performance. If investment responds positively to net performance, the reduced performance counteracts further circular efficiency investment. If investment responds differently, the feedback sign and strength must be reassessed empirically rather than assumed.

5.3. Integrated Causal Loop Diagram

Figure 2 reports seven selected closed feedback paths derived from the causal relationships in Section 4 and Section 5, with an exact matching path for each entry in Table 7. R1–R3 describe recovery investment, organizational learning, and quality-driven market development. B1–B4 describe operating burden, delayed processing-capacity adjustment, secondary-market price pressure, and demand-mediated rebound. Each path is shown with its return link rather than an unclosed causal chain; links are hypothesized and can share variables across paths. The complete model is intended as a qualitative explanation, not a numerical simulation or a demonstration of loop dominance.

The model shows why circular supply chain performance cannot be understood from recovery rates alone. When R1–R3 dominate, circular intervention strengthens capability, material quality, secondary demand, investment, and sustainable value. When B1–B4 become consequential, operating burden, delayed capacity response, secondary market imbalance and rebound demand can constrain net returns under their stated conditions. The resulting sustainability performance therefore depends on the relative strength of reinforcing value-creation processes and counteracting value-leakage processes rather than on circular intervention intensity alone.

6. System Conditions and Contextual Evidence

The causal structure suggests that circular interventions can produce different sustainability outcomes depending on the alignment among recovery activity, organizational capability, processing capacity, secondary markets, and demand responses. Four system conditions are therefore considered: fragmented circularity, scale without alignment, coordinated circularity, and value-preserving circularity. These conditions are qualitative representations rather than forecasts.

6.1. Alternative System Conditions

Under fragmented circularity, individual reuse, recycling, recovery, or reverse logistics initiatives exist, but coordination across the supply chain remains limited. Some materials are successfully recovered, yet weak information exchange, inconsistent quality, limited processing capacity, and uncertain markets create substantial value leakage. Increasing circular intervention can improve performance initially, but benefits remain constrained by the surrounding system. A more problematic condition occurs when circular activities are expanded primarily through collection or recovery targets without corresponding development of infrastructure, organizational capability, and secondary markets. This scale-without-alignment condition increases reverse flow volume while processing and market capacity remain relatively fixed. The balancing loops identified in Section 5 become stronger: congestion increases, recovery efficiency declines, secondary material inventories grow, and operating costs rise. Circularity indicators may improve while net sustainability gains increase only marginally or potentially deteriorate.

Coordinated circularity occurs when recovery expansion is accompanied by investment in processing capacity, information systems, organizational capability, stakeholder coordination, and secondary markets. Under this condition, the reinforcing loops are more likely to dominate. Recovery improves material availability, higher-quality outputs strengthen secondary demand, and economic returns support further capability development. The objective is therefore not maximum circular intervention but alignment between intervention and system capacity. The fourth condition extends this logic by combining efficient recovery with waste prevention, product-life extension, demand management, and higher-value retention strategies. Value-preserving circularity prioritizes reduction, reuse, repair, and remanufacturing where feasible before relying on material recycling and lower-value recovery. It also limits rebound pressure by focusing on absolute resource use and total system throughput rather than efficiency per unit alone [8,16,20]. Table 8 summarizes these alternative system conditions and their expected implications for circular supply chain development.

Table 8. Alternative system conditions for circular supply chain development.

System Condition	Circular Intervention	Capability and Coordination	Recovery and Market Capacity	Expected System Behavior	References
Fragmented circularity	Low–moderate	Uneven	Limited or fragmented	Some recovery gains but substantial value leakage	Govindan and Hasanagic [11]; Farooque et al. [6]
Scale without alignment	High and rapidly expanding	Develops slowly	Capacity and markets lag behind recovery	Congestion, higher cost, declining marginal benefit, market pressure	Govindan and Soleimani [7]; Zink and Geyer [8]; Shamsuddoha et al. [36]
Coordinated circularity	Moderate–high	Strong	Develops with circular flows	Recovery, quality, investment, and market demand reinforce one another	de Angelis et al. [30]; Farooque et al. [6]; Velenturf and Purnell [20]
Value-preserving circularity	Selective and system-oriented	Strong	Aligned with resource hierarchy and demand	Higher value retention with lower value leakage and rebound pressure	Korhonen et al. [16]; Castro et al. [10]

Note: Conditions represent conceptual ideal types; the table is not a historical classification or a forecast.

Table 8 illustrates that increasing circular intervention does not lead to a single predetermined outcome. The expected system behavior depends on whether capability, coordination, recovery capacity, and market capacity develop sufficiently to support the scale of circular activity.

6.2. Illustrative Evidence from the United States and Bangladesh

The United States and Bangladesh provide useful contrasting contexts for interpreting these system relationships. They are not used as matched cases, and differences in definitions, institutional structures, waste systems, and data availability prevent direct quantitative comparison. In the United States, circular economy development occurs within comparatively mature logistics and industrial systems, yet important structural constraints remain. The U.S. Environmental Protection Agency’s National Recycling Strategy identifies reduced markets for recycled materials, inadequate infrastructure, contamination, consumer confusion, and inconsistent performance measurement as continuing challenges. The strategy specifically links stronger recycling performance with improved secondary markets, collection and infrastructure, reduced contamination, supportive policy, and better measurement [48]. These conditions closely correspond with the capacity and market feedback mechanisms represented in the proposed model.

Mechanism-to-evidence trace. The EPA report identifies constrained processing infrastructure and contaminated recovery streams (relevant to B2 and P4), reduced markets for recovered commodities (B3 and P6), and explicit attention to equitable access and worker safety (P9) [48]. These are documentary observations and policy priorities, not measured coefficients or confirmation of the proposed causal-loop closure. They specify plausible conditions under which the proposed capacity, market and social relationships warrant testing.

The World Bank documents increasing plastics use and limited formal recycling in Dhaka and identifies infrastructure, technology, stakeholder participation, green skills and jobs as policy priorities [49]. These observations make return-volume pressure, capacity alignment (B2; P4 and P10), and social inclusion (P9) relevant hypotheses. They do not independently demonstrate that the model’s predicted feedback relationships occurred or that U.S. and Bangladeshi systems are statistically comparable.

The U.S. case also demonstrates why recycling alone should not be equated with circularity. EPA explicitly recognizes that municipal solid waste recycling represents only one component of a broader circular economy and that source reduction, reuse, product design, and other material strategies are required alongside recycling [48]. This supports the distinction made in the framework between increasing circular output and generating sustainable value. Bangladesh illustrates different but related system constraints. Rapid urbanization and increasing consumption have intensified pressure on material recovery systems. World Bank evidence indicates that annual per capita plastic consumption in urban Bangladesh increased from approximately 3 kg in 2005 to 9 kg in 2020, while consumption in Dhaka reached approximately 22.5 kg per capita. The same evidence reported that only 37.2% of plastic waste in Dhaka was recycled, illustrating the gap that can emerge between growing material throughput and available recovery systems [49]. Bangladesh’s National Action Plan for Sustainable Plastic Management adopts a circular approach based on reduction, reuse, and recycling and identifies infrastructure, technology, institutional capacity, investment, behavioral change, and stakeholder participation as interconnected requirements. The plan established phased targets extending to 2030 and emphasized that circular material flows require interventions across production, consumption, waste management, and regeneration rather than recycling alone [49]. These conditions reinforce the system argument developed here: increasing recovery targets without parallel development of infrastructure, markets, institutional capability, and stakeholder coordination can strengthen balancing constraints. Table 9 summarizes the contextual evidence from the United States and Bangladesh and relates it to the mechanisms represented in the proposed framework.

Table 9. Contextual evidence and implications for the circularity paradox.

Context	Observed System Issue	Relevance to the Proposed Framework	References
United States	Recycling infrastructure has not consistently kept pace with a diverse and changing waste stream	Supports the recovery-capacity constraint	U.S. EPA [48]
United States	Reduced and uncertain markets for recycled materials remain a recycling challenge	Supports secondary market absorptive capacity	U.S. EPA [48]
United States	Contamination and inconsistent recycling practices affect material recovery	Supports recovered-quality and value-leakage mechanisms	U.S. EPA [48]
United States	Recycling alone is insufficient for achieving a circular economy	Supports distinction between circular activity and sustainable circular performance	U.S. EPA [48]
Bangladesh	Urban plastic consumption increased substantially between 2005 and 2020	Illustrates throughput pressure on recovery systems	World Bank [49]
Bangladesh	Only 37.2% of plastic waste in Dhaka was reported as recycled in the cited 2021 evidence	Illustrates the gap between material generation and recovery	World Bank [49]
Bangladesh	Sustainable plastic management requires infrastructure, technology, investment, institutional capacity, and stakeholder coordination	Supports capability-alignment relationships	World Bank [49]
Bangladesh	National strategy combines reduction, reuse, recycling, design, waste management, and regeneration	Supports system-level rather than recycling-only circularity	World Bank [49]
United States	National Recycling Strategy specifies worker health, equitable participation and access as program concerns.	Grounds proposed social-outcome indicators and conditional P9; not observed effect.	U.S. EPA [48]
Bangladesh	Policy report connects circular plastics to possible green skills and employment.	Supports testing P9 with worker and inclusion indicators, not a claim of delivered job outcomes.	World Bank [49]

Note: Sources document reported conditions or policy priorities. Corresponding loops are interpretations to be tested, not confirmed effects.

The two contexts therefore illustrate a common theoretical point despite substantial institutional differences. Circularity becomes constrained when material flows expand faster than the systems required to recover, process, coordinate, and absorb them. The location of the bottleneck may differ, but the underlying feedback structure remains relevant.

7. Discussion

The proposed framework challenges a simple linear assumption frequently associated with circular supply chain implementation:

More circular activity → more resource recovery → greater sustainability. 

The system dynamics perspective suggests a more conditional relationship. Circular intervention increases recovery opportunities, but it also changes the volume and complexity of reverse flows. As these flows expand, organizational, logistical, processing, and market requirements increase. Sustainable performance therefore depends on whether supporting capabilities develop sufficiently to prevent these requirements from becoming sources of value leakage.

7.1. Explaining the Circularity Paradox

The paradox emerges from the simultaneous operation of reinforcing and balancing feedback. R1–R3 encourage circular expansion through recovery value, capability development, and secondary market development. These processes explain why successful circular initiatives can become increasingly viable as experience, scale, infrastructure, and market confidence develop. The balancing loops explain why the same expansion does not continue indefinitely without additional system adjustment. Reverse flow complexity increases management requirements; capacity constraints create congestion; growing secondary supply can exceed market absorption; and efficiency gains can stimulate rebound demand. The sustainability outcome depends on the relative strength of these opposing processes. The framework therefore does not identify a universal level of optimal circularity, which is consistent with recent research emphasizing contextual barriers, trade-offs, and system-specific transition conditions [50,53,58]. A recovery rate that is sustainable in one supply chain may generate high cost, or environmental burden in another because transportation distance, material value, processing technology, energy sources, market demand, or organizational capability differ. Similarly, a circular intervention that initially produces strong benefits may experience diminishing returns as scale increases.

7.2. Circularity as a Capability–Capacity Alignment Problem

An important implication is that circular supply chain development can be understood as an alignment problem. Circular intervention creates demand for organizational capability, logistics capacity, recovery infrastructure, stakeholder coordination, and secondary market absorption. Sustainable value is more likely to increase when these supporting elements develop together. This interpretation extends dynamic capability thinking into circular supply chain systems. Organizational capability does not merely improve performance directly; it changes the level of circular complexity that a supply chain can manage before balancing pressures dominate. Stronger capability can therefore shift the point at which circular intervention begins experiencing diminishing returns. The same principle applies to infrastructure and markets. Expanding collection without processing capacity moves the bottleneck downstream. Expanding processing without secondary demand creates recovered material inventories. Expanding secondary demand without reliable material quality creates supply uncertainty. Circular supply chain performance consequently depends on coordinated development across the complete system.

7.3. Propositions for Future Empirical Testing

The causal framework produces several propositions that can be examined through future survey research, case analysis, longitudinal data, or quantitative system dynamics simulation. Table 10 presents eleven propositions derived from the reinforcing and balancing relationships identified in the framework.

Table 10. Theoretical propositions derived from the circularity paradox framework.

Proposition	Expected Relationship	References
P1	P1. At a fixed product and system boundary, increasing circular intervention intensity is expected to improve absolute virgin-material displacement and recovery value when processing capacity and secondary demand are sufficient; otherwise, marginal gains decline.	Genovese et al. [4]; Farooque et al. [6]
P2	P2. As reverse-flow heterogeneity and distance rise relative to organizational capability, the positive association between circular intervention intensity and net environmental and economic performance is expected to weaken through transport, sorting and delay burdens.	Govindan and Soleimani [7]; de Angelis et al. [30]
P3	P3. Higher organizational capability, proxied by traceability, quality decision performance and process learning, is expected to strengthen the positive effect of circular activity on recovery yield, controlling for the mix and condition of returns.	Teece et al. [28]; Eisenhardt and Martin [29]; de Angelis et al. [30]
P4	P4. When arrivals of returned material persistently exceed effective processing completions, unprocessed-return inventory and congestion are expected to rise and per-unit recovery yield and environmental net benefit to decline, holding material composition constant.	Shamsuddoha et al. [36]; Shamsuddoha et al. [37]
P5	P5. Greater verified secondary demand and suitable output quality are expected to strengthen the effect of recovery volume on virgin-material displacement and recovered sales value, conditional on genuine substitution for primary materials.	Zink and Geyer [8]; Velenturf and Purnell [20]
P6	P6. When recoverable supply persistently exceeds market uptake, the increasing supply–demand imbalance is expected to lower realized recovered-material prices and, through profitability, subsequently reduce investment in recovery capacity, other things equal.	Zink and Geyer [8]; Velenturf and Purnell [20]
P7	P7. Where lower effective resource or product costs increase consumption, stronger rebound demand is expected to reduce absolute material and emissions savings despite improvements in circular efficiency per unit; the effect depends on actual primary-product substitution.	Makov and Font Vivanco [9]; Castro et al. [10]
P8	P8. Simultaneous improvements in capability, processing availability, stakeholder coordination and market uptake are expected to produce higher environmental and economic circular performance than comparable increases in collection activity without such alignment.	Farooque et al. [6]; Castro et al. [10]
P9	Under otherwise comparable levels of circular collection and recovery activity, programs that include enforceable worker safety standards, meaningful community participation, and equitable inclusion of informal recovery actors are expected to produce higher social circular performance than programs lacking such safeguards; candidate indicators include occupational injury rates, job quality, access to formal recovery markets, participation in relevant decisions, and the distribution of local environmental burdens.	U.S. EPA [48]; World Bank [49]
P10	When new recovery capacity has a longer commissioning delay than the growth of return arrivals, unprocessed-return backlog and congestion-related value leakage are expected to peak before completed capacity becomes available; the timing and size of this effect depend on initial slack capacity.	Sterman [15]; Shamsuddoha et al. [36,37]
P11	Where consumer and producer adoption responds with a lag to lower effective costs from circular efficiency, rebound demand and total material throughput are expected to rise after the initial per-unit efficiency benefit, weakening subsequent absolute environmental savings.	Zink and Geyer [8]; Makov and Font Vivanco [9]; Castro et al. [10]

Note: P1–P11 are conditional directional propositions with candidate indicators, not statistically supported causal estimates.

Together, these eleven conditional propositions translate the conceptual framework into relationships with candidate indicators, mediating mechanisms, boundary conditions, and directions for future empirical tests. They provide a basis for examining how capability, capacity, market absorption, and rebound conditions influence the relationship between circular intervention and sustainable circular performance.

For operational testing, P1–P8 can be assessed using plant-level or supply-chain panel data on recovered mass, contamination, process capacity and throughput, unit costs, realized secondary sales, prices, and verified displacement of primary material. P9 requires a separate social measurement protocol addressing worker safety, job quality, fairness and community access, without converting these outcomes into unweighted monetary equivalents. P10–P11 require longitudinal observations of commissioning time, returns backlog, price or accessibility changes and delayed consumption, while controlling for exogenous market movements.

8. Theoretical Contributions

The study makes four principal theoretical contributions. First, it separates circular intervention intensity from sustainable circular performance. Circular supply chain studies frequently use recovery, recycling, reuse, or circular practice adoption as indicators of progress. The present framework argues that these measures describe circular activity but do not independently establish net sustainability performance. Sustainable circular performance requires consideration of the value retained and the environmental and economic burdens generated in achieving that retention.

The more precise theoretical distinction is between operational leakage that arises inside the recovery system even without a demand response, and rebound that arises through induced consumption or production responses to circular interventions. Both can reduce the environmental dimension of net circular performance, but they require different mediators, indicators and policies. By keeping these pathways analytically separate and linking them through conditional feedback and delays, the study contributes a causal integration rather than renaming well-established rebound mechanisms.

Second, the study introduces value leakage as a connecting mechanism between circular supply chain operations and circular rebound. Existing rebound research has established that efficiency and circular business models can stimulate behavioral and economic responses that offset environmental gains [8,9,10]. The proposed framework extends this reasoning by incorporating operational leakage through reverse logistics complexity, processing burden, capacity pressure, degradation, and secondary market constraints. Circular benefits can therefore be eroded through operational mechanisms even when a conventional consumption rebound is limited.

Third, the study makes social outcomes analytically visible alongside economic and environmental effects. It specifies worker protection, inclusion in recovered-material markets, community participation, and equitable exposure to processing burdens as outcomes to be independently measured, rather than implying that improved recovery automatically improves social welfare.

Fourth, the framework integrates dynamic capabilities with system dynamics. Organizational capability affects the ability of the system to absorb increasing circular complexity, while feedback from successful recovery can itself strengthen capability through learning and investment. Capability is consequently both a determinant and an endogenous outcome of circular development. This provides a dynamic explanation of why similar circular interventions can produce different outcomes across firms, industries, and institutional environments.

9. Managerial and Policy Implications

For managers, the framework suggests that circular supply chain strategies should not be evaluated primarily through the quantity collected, returned, or recycled. Performance systems should also monitor recovery yield, transportation intensity, processing cost, recovered material quality, inventory accumulation, secondary market demand, virgin material displacement, and the net value retained through recovery. Increasing collection while recovered inventories accumulate or processing costs rise may indicate that the system is approaching a capacity or market constraint. Investment decisions should similarly be coordinated across the circular supply chain. Expanding reverse logistics requires corresponding consideration of sorting and processing capacity, information systems, recovery technology, workforce capability, and downstream markets. Digital visibility can improve coordination, but information alone cannot compensate for inadequate physical capacity or weak secondary demand. For policymakers, circular targets should avoid encouraging one-dimensional optimization. Collection or recycling targets can increase circular activity while transferring pressure to sorting facilities, recyclers, secondary markets, or other parts of the system. Policies combining waste prevention, design for recovery, collection infrastructure, recycled-content demand, market development, standards, information, and capacity building are more consistent with the system relationships identified in this study. The U.S. National Recycling Strategy similarly links stronger recycling with market development, infrastructure, contamination reduction, policy coordination, and improved measurement rather than relying on collection alone. Bangladesh’s plastic management framework likewise emphasizes coordinated action across design, consumption, waste management, regeneration, infrastructure, technology, and institutional capacity.

A multidimensional monitoring system should report social performance alongside technical indicators. Managers and policymakers should track collection and sorting injury risks, job quality and training, participation of affected communities and informal workers, and whether environmental burdens move toward already exposed communities. Targets for higher recovered tonnage should be evaluated against these outcomes rather than treated as sufficient evidence of inclusive sustainability [48,49].

10. Limitations and Future Research

10.1. Limitations

This study develops a qualitative theoretical framework and therefore has several limitations. The causal relationships are grounded in the literature and secondary evidence but are not estimated statistically. The causal loop model identifies expected directions and feedback mechanisms rather than numerical effect sizes, thresholds, or forecasts. The proposed circularity threshold should consequently be interpreted as a theoretical system condition rather than a predetermined numerical point. The United States and Bangladesh evidence is illustrative rather than comparative. Differences in economic structure, waste definitions, regulatory systems, informal sector participation, infrastructure, and data quality prevent direct comparison. Future research could examine the framework within individual industries or countries using standardized longitudinal data.

A further methodological reporting limitation is that no contemporaneous, reproducible historical literature-search log was provided for this revision. The retrospective selection and causal coding rules explain how the present model was reconstructed but do not establish exhaustive coverage. The structural audit of the causal diagram is an internal consistency check; it is not independent expert verification or empirical behavioral validation. Neither the seven paths nor the propositions have been tested with primary observations. Finally, social, economic and environmental dimensions have not been combined using validated weights or a common functional-unit framework, and the diagram omits numerical stock-and-flow equations. These limitations qualify the strength of the paper’s inference.

10.2. Priorities for Future Research

The most important extension would be development of a quantitative stock-and-flow system dynamics model. Stocks could include accumulated returned products, recoverable materials, recovery capacity, secondary material inventory, organizational capability, and circular investment. Flows could represent product returns, processing, material recovery, secondary material consumption, disposal, capacity investment, and capability development. Parameter estimation would permit sensitivity analysis and identification of conditions under which reinforcing or balancing loops become dominant. Future empirical studies could also test the propositions through survey data, longitudinal case studies, structural equation modeling, or multi-industry datasets. Sector-specific applications would be particularly valuable because circularity constraints vary considerably among plastics, electronics, automotive products, textiles, food, construction materials, and agricultural or livestock systems.

Future studies should therefore proceed in stages: document a reproducible database search and screening record; independently review link definitions with domain experts and affected workers; estimate observed stocks and flows within defined industry and geographic boundaries; and test loop directions and delays with longitudinal operational data. Quantitative models should test backlog accumulation, capacity commissioning delays, market response and rebound under alternative policy assumptions. Environmental life-cycle measures, economic accounts and social indicators should initially remain separate, with any later composite index justified transparently to affected stakeholders. Context-specific testing in the United States, Bangladesh or another well-documented setting would be preferable to assuming direct cross-country equivalence.

11. Conclusions

Circular supply chains offer important opportunities to reduce waste, retain resource value, and decrease dependence on virgin materials. However, increasing circular activity does not automatically generate proportional sustainability improvement. Circular interventions simultaneously create recovery opportunities and additional logistical, processing, organizational, market, and behavioral requirements. This study conceptualizes this tension as the circularity paradox and develops a qualitative system dynamics framework explaining its underlying feedback structure. Three reinforcing mechanisms support circular development through recovery value, capability improvement, and secondary market development. Four balancing mechanisms constrain these gains through reverse flow complexity, processing capacity pressure, secondary market saturation, and rebound demand. The central implication is that sustainable circularity depends on system alignment rather than circular intensity alone. Circular intervention must develop together with organizational capability, stakeholder coordination, recovery infrastructure, material quality, and secondary market demand. When these elements remain aligned, circular activity can reinforce sustainable value creation. When circular activity expands faster than the capacity of the surrounding system, value leakage and rebound effects can progressively reduce its marginal sustainability benefit. Circularity should therefore be viewed not as an end state but as a means of preserving economic, environmental, and social value. The relevant question for sustainable supply chain management is not simply how much material can be circulated, but how much sustainable value remains after the complete system required to circulate that material is considered.

The study’s contribution is conceptual: the loop signs have been internally checked, but neither the seven paths nor the proposed performance effects constitute causal estimates. Future research must test their boundary conditions, dynamic delays, and economic, environmental and social outcomes with appropriate observational, stakeholder and simulation evidence.