Unpacking the underlying mechanisms between blockchain technology application and green innovation: from systemic resource orchestration perspective

1 Introduction

There are growing concerns about the environmentally unethical behavior of enterprises worldwide. In China, which is a major manufacturing country, the value added of manufacturing accounts for approximately 30% of the global total and has ranked first worldwide for 14 consecutive years. In recent years, manufacturing has been confronted with severe environmental problems and the need for economic transformation. However, as a result of the characteristics of high energy consumption and high emissions inherent in manufacturing enterprises, these enterprises inevitably bear important responsibilities in the green transformation of the economy and society (Costa and Opare, 2025; Tjahjadi et al., 2020). Therefore, the Chinese government not only promulgated the Environmental Protection Law in 2015, setting higher environmental criteria and stricter regulatory actions for enterprises that cause serious environmental damage (such as the steel, petrochemical, and power industries), but also proposed carbon peaking and carbon neutrality goals (Coussa et al., 2025). As a result, environmental protection has become an inevitable “investment” for manufacturing enterprises. Most enterprises have come to realize that passive end-of-pipe treatment strategies are insufficient to cope with increasing institutional and environmental pressures; thus, they need to proactively carry out green innovation and achieve sustainable development (Li et al., 2025). However, green innovation and government environmental protection regulations significantly increase the production and operating costs of enterprises, further stretching the already strained resources of enterprises.

From a systemic perspective, this dilemma is not merely a linear problem of increasing costs; rather, it is a complex challenge arising from the dynamic interplay among institutional pressures, technological upgrading needs, and internal resource constraints (Kast and Rosenzweig, 1972). This systemic view requires managers to reconceptualize their enterprise not as an isolated linear machine but as an open complex system where environmental strategy, operational processes, and resource allocation are deeply interconnected with the external environment (Scott and Davis, 2015). To address such complexity, this study integrates a systemic perspective with resource orchestration theory. Resource orchestration theory traditionally explains how firms actively manage resource portfolios within their own boundaries, but it does not fully account for cross-boundary resource dynamics. By adding a systemic lens, the combined framework elevates the unit of analysis from the individual firm to the entire supply chain system, revealing how resource structuring, bundling, and leveraging operate across organizational boundaries. For enterprises facing resource scarcity, the new challenge is no longer just “how much” to invest; rather, it is how limited resources can be effectively orchestrated across the supply chain system to promote green innovation.

Among various digital technologies, blockchain stands out as a particularly promising infrastructure for green innovation (Alexopoulos et al., 2022; Chen, 2025). Blockchain technology is an emerging technology that integrates distributed ledgers, smart contracts, and other technologies. It is adept at information tracing, information sharing, and self-execution (Ali et al., 2025; Dubey et al., 2020) as well as optimizing operational processes (Ahmed et al., 2022) and gathering new ideas (Giri and Mannhar, 2023), all of which are conducive to green innovation. However, many previous studies have been limited to exploring the direct effect of blockchain technology application (BTA) on green innovation (i.e., El-Kassar and Singh, 2019; Luo et al., 2023; Vaia et al., 2022) and some studies identified mediators (such as supply chain coordination and supply chain integration) without delving into the specific mechanisms (i.e., Jie et al., 2024; Razak et al., 2025; Zakaria et al., 2024). In adopting a systemic perspective, green innovation should be conceptualized not only as a firm-level capability that emerges in isolation but also as an emergent property of the entire supply chain system through resource orchestration (Chen and Cui, 2022; Dong et al., 2025). BTA reshapes the structural and dynamic characteristics of supply chain systems, enabling systemic integration that transcends organizational boundaries. Enterprises acquire brand-new key knowledge from the outside more actively (Duan et al., 2021), and supply chain partners, which constitute the core of the external relationships of enterprises, can provide scarce knowledge and resources for enterprises through supply chain integration (Chu et al., 2019). Enterprises can reduce the uncertainty risks faced in the process of green innovation, fill gaps in knowledge about green innovation, and improve the efficiency of green innovation by integrating with supply chain partners. Therefore, drawing on systems thinking and resource orchestration theory, this study conceptualizes supply chain integration as a critical system-level resource-bundling mechanism that transforms structured digital resources into systemic green innovation capabilities.

However, applying blockchain technology for green innovation does not always lead to success (Najati, 2025). Some scholars have found that blockchain technology gives each block/partner an equal status, which leads to decentralized decision-making and low efficiency in the implementation of innovative activities (Zhan et al., 2025). Meanwhile, the application of new technologies primarily reduces green innovation efficiency by increasing the pressure for R&D cost savings (Zhao and Li, 2025). Thus, it is necessary to concern about contexts that affect the impact of BTA on green innovation. These specific system-level contingencies create nonlinear feedback loops (supply chain trust as an amplifier) or structural constraints (task complexity as a bottleneck), while the cultural-cognitive orientation (green digital learning orientation) determines the direction of resource leveraging.

The transformation of digital resources into integrated capabilities is contingent upon specific contextual conditions (Ahmed et al., 2022). The technical and structural characteristics of blockchain restrict the transformation conditions from structured digital resources to bundled resources; that is, they limit the efficiency of BTA for supply chain integration. Processing transaction information in blockchain requires supply chain partners to have a fully trusting cooperative atmosphere and to not restrict resource sharing because of concerns about sensitive information leakage or opportunistic behavior (Dubey et al., 2020). Only in this way can a comprehensive and in-depth digital resource-sharing situation be achieved and the goal of effectively enhancing supply chain integration through BTA be reached. Meanwhile, the information and data shared in blockchain must be encodable (Ante, 2020). Thus, transaction tasks and cooperation performance must be encoded into computer codes for operation. However, if the tasks are essentially difficult to convert into computer languages for encoding, the efficiency of BTA for supply chain integration will decrease (Lumineau et al., 2021). Therefore, it is necessary to verify the moderating effects of supply chain trust and task complexity on the relationship between BTA and supply chain integration.

The strategic priority guiding the bundling of resources determines whether integrated resources can be effectively transformed into advanced capabilities (Chen and Cui, 2022). From a systemic perspective, a green digital learning orientation serves as the cultural-cognitive subsystem that sets the goal-direction for leveraging bundled resources. It reflects the extent to which the supply chain system collectively prioritizes environmental sustainability as a learning objective within digitally mediated collaboration, actively seeking, creating, and applying green knowledge through digital technologies (Benzidia et al., 2021; Halbusi et al., 2024). Such an orientation in enterprises is likely to lead firms to invest their integrated resources in environmentally friendly technologies and to promote ethical business processes to lower their carbon footprint (Wan et al., 2020). Therefore, it is important to examine the moderating role of a green digital learning orientation in the relationship between supply chain integration and green innovation.

To that end, applying a systemic perspective and resource orchestration theory, this study explores the effect of BTA on green innovation and the mediating role of supply chain integration, as well as the moderating effects of supply chain trust, task complexity, and green digital learning orientation. By doing so, it extends the discourse on digital business value creation beyond traditional firm boundaries to include complex digital ecosystems within supply chains. Importantly, while recent empirical studies have confirmed that supply chain integration mediates the relationship between digital technology adoption and green innovation (e.g., Zakaria, Ming and Hashim, 2024; Jie, Hui Xin and Rahmat, 2024), they have focused primarily on whether integration mediates rather than why and how it does so from a theoretical standpoint. Departing from prior research that treats green innovation as a firm-level outcome driven by environmental regulations or government subsidies (Coussa et al., 2025; Guo et al., 2023; Li et al., 2025), this study draws on resource orchestration theory and a systemic perspective to unpack the systemic orchestration mechanisms that explain why integration functions as a bundling mechanism. Specifically, this study theorizes that BTA accomplishes system-level structuring, supply chain integration accomplishes system-level bundling through three functions (complementary resource combination, systemic friction reduction, and dynamic responsiveness), and green innovation accomplishes system-level leveraging. Moreover, this study simultaneously examines three system-level moderators (supply chain trust as a relational amplifier, task complexity as a structural constraint, and green digital learning orientation as a cultural-cognitive goal-setting mechanism) that prior studies have treated in isolation or not at all. By revealing nonlinear feedback loops (trust amplifies BTA’s effect), system boundary conditions (task complexity as a structural constraint), and the cultural-cognitive directing role of green digital learning orientation, this study reconciles conflicting findings on the BTA-green innovation relationship (cf. Zhan et al., 2025; Zhao and Li, 2025). Thus, rather than merely documenting mediation, this integrated framework unpacks the systemic orchestration mechanisms linking BTA to green innovation from a systemic resource orchestration perspective. The conceptual model of this study is shown in Figure 1.

2 Literature review

2.1 Systemic perspective and resource orchestration theory

A systemic perspective views an enterprise not as an isolated entity but as an open complex system embedded within a broader network of interdependencies, including supply chain partners, institutional environments, and technological ecosystems (Scott and Davis, 2015). This perspective also emphasizes that understanding how firms achieve competitive advantage requires examining the structural relationships, feedback loops, and coordination mechanisms that shape resource transformation processes across organizational boundaries (Madlenak et al., 2025). Thus, organizational capabilities, such as green innovation, are not merely the sum of individual resources; rather, they are emergent properties arising from the dynamic interactions and resource flows among system elements (Kast and Rosenzweig, 1972; Scott and Davis, 2015).

Resource orchestration theory provides a theoretical framework for understanding how firms actively manage and transform resources within these complex systems. Resource orchestration theory posits that an enterprise’s superior capability is crucial for achieving and maintaining competitive advantage, emphasizing that capability growth hinges not only on resource acquisition and accumulation but also predominantly on resource orchestration (Sirmon et al., 2011; Dessaigne and Pardo, 2020). Enterprises must strategically allocate their resources to gain a competitive edge in managing unforeseen environmental and corporate challenges (Chen and Cui, 2022). This theory delineates the capability development process, which comprises three components: the structuring of a resource portfolio (which entails acquiring, accumulating, and divesting resources to establish the firm’s resource portfolio), the bundling of resources through integration and combination (which focuses on integrating resources to create capabilities), and the leveraging of configurations to capitalize on market opportunities based on the firm’s capabilities (which encompasses formulating a plan or vision for capabilities, integrating capability configurations, and implementing strategies) (Sirmon et al., 2011). The three processes are especially beneficial for understanding resource allocation and capability enhancement (Chen and Cui, 2022; Liu et al., 2016).

This study articulates how the systemic perspective extends resource orchestration theory beyond firm-level boundaries. Resource orchestration theory traditionally answers the question of “how firms actively manage resource portfolios” to build capabilities and achieve competitive advantage, but it largely assumes that structuring, bundling, and leveraging occur within a single organizational boundary (Sirmon et al., 2011; Chen and Cui, 2022). By contrast, a systemic perspective elevates the unit of analysis to the supply chain system, introducing distinctive theoretical advantages that are absent from firm-level resource orchestration accounts. First, emergence means that green innovation is not simply the sum of individual firms’ green investments or capabilities; rather, it emerges from coordinated interactions, information flows, and joint problem-solving across multiple actors in the supply chain system (Kast and Rosenzweig, 1972; Mitleton-Kelly, 2003). Second, nonlinear feedback loops imply that relational factors such as supply chain trust do not have a constant linear effect (Kauffman, 1993; Scott and Davis, 2015); instead, they can amplify or dampen the impact of digital technologies such as BTA through reciprocal, system-wide interactions. Third, system boundary conditions refer to structural features of the interorganizational system (Mitleton-Kelly, 2003), like task complexity, which act as constraints that determine how easily digital resources can be codified, shared, and integrated across partners. Fourth, the systemic perspective adds a cultural-cognitive goal-setting element (Scott and Davis, 2015). This refers to shared orientations, for example, a green digital learning orientation, that function as the system’s internal compass, directing the leveraging of bundled resources toward sustainability objectives rather than competing priorities. Integrating these systemic concepts with resource orchestration theory, this study reconceptualizes the three orchestration processes as system-level mechanisms. Structuring depends on blockchain’s systemic transparency to identify and allocate distributed resources across the supply chain network. Bundling relies on supply chain integration as a system-wide coordination mechanism that combines complementary resources and reduces systemic friction. Leveraging requires a green digital learning orientation as a cultural-cognitive subsystem that sets goal-direction for transforming bundled resources into green innovation capabilities. The integrated framework not only preserves the processual logic of resource orchestration but also enriches it with system-level emergence, feedback dynamics, contextual constraints, and shared cognitive guidance.

2.2 Green innovation and BTA

Green innovation takes into account both economic and environmental benefits, regards ecological priority as the principle, and strives to achieve the goals of clean production, energy conservation, emission reduction, and a low-carbon cycle (Chen et al., 2006; Dong et al., 2025). From the systemic perspective and resource orchestration view, green innovation is an advanced capability that emerges from the dynamic orchestration of resources across the supply chain system. Enterprises need to constantly integrate internal and external resources and master mainstream technologies, new concepts, and development trends in the process of achieving green innovation (Teodorescu and Korchagina, 2021). Within this systemic process, BTA functions as a digital infrastructure that reconfigures information flows, knowledge networks, and coordination mechanisms across organizational boundaries (Benzidia et al., 2021; Lumineau et al., 2021). Thus, blockchain technology can help enterprises effectively carry out green innovation.

Previous studies have explored the effect of BTA on green innovation from different perspectives. First, from the perspective of information integration and sharing, in contrast to traditional innovation, green innovation involves highly comprehensive information, such as energy conservation, pollution prevention and control, waste utilization, and clean production (Vaia et al., 2022). From a systemic view, BTA reduces systemic uncertainties by enabling real-time, transparent, and tamper-proof information flows across the supply chain network (Dubey et al., 2020; Luo et al., 2023). It can break down information barriers; achieve the collection, integration and sharing of information related to green, low-carbon and environmental protection; help enterprises fully grasp the information needed for green innovation; and provide technical support for the smooth transmission and efficient sharing of information, ultimately reducing the risks of green innovation and promoting the green innovation of enterprises (El-Kassar and Singh, 2019). Second, from the perspective of knowledge integration, green innovation, which is an innovative activity in interdisciplinary studies, encompasses knowledge from multiple technical fields. BTA expands a firm’s knowledge search space by connecting it to a broader network of external knowledge sources, thereby enhancing the system’s capacity for knowledge recombination and novelty generation (Ahmed et al., 2022) and leading to more cutting-edge green innovations. Finally, from the perspective of property rights protection, BTA can shape a fair technological ecosystem through encryption algorithms and provide a better rights protection environment for green innovation, thereby promoting green innovation (Luo et al., 2023).

However, the negative effects of BTA on green innovation cannot be ignored (Najati, 2025). These negative effects can be understood as unintended consequences arising from the structural characteristics of blockchain technology interacting with existing system dynamics. The high energy consumption problem of blockchain technology itself also poses challenges to green innovation (Barron et al., 2021). In addition, blockchain technology gives each partner an equal status, which decentralizes decision-making and leads to low efficiency in the implementation of innovation activities (Zhan et al., 2025). Meanwhile, the adoption of high and new technologies also reduces the efficiency of green innovation by increasing the pressure to save on R&D costs (Zhao and Li, 2025). Therefore, these opposing views emphasize the necessity of in-depth analysis when exploring the relationship between BTA and green innovation. From an integrated systemic perspective and resource orchestration view, this study explores the mediating and moderating mechanisms through which BTA influences green innovation, with the aim of reconciling the conflicting results of previous research by uncovering the systemic conditions under which BTA facilitates or hinders green innovation outcomes.

2.3 The vital role of supply chain integration

Because of the complexity, uncertainty, and high cost of the green innovation process, an increasing number of enterprises are actively seeking external resources that complement their internal innovation resources by enhancing integration with supply chain partners to reduce the uncertainty risks faced in the innovation process and compensate for their resource gap (Elysé et al., 2025; Gauthier and Genet, 2014). From a systemic perspective, the supply chain is conceptualized as a complex adaptive system, wherein supply chain integration serves as a key mechanism for coordinating interdependencies and enabling systemic resource flows (Chu et al., 2019). Supply chain integration is widely regarded as a state of close collaboration across organizational boundaries, including coordinating functions, exchanging information, and sharing logistics flows with key suppliers and customers (Kim and Schoenherr, 2018; Swink et al., 2007). The core objective of supply chain integration is to break the “silo effect”, achieve an overall collaborative optimization of the supply chain, and enhance efficiency, response speed, and competitiveness (Ali et al., 2025).

Recent empirical work has provided valuable evidence that supply chain integration (or collaboration) mediates the relationship between blockchain application and green innovation outcomes. For instance, Razak et al. (2025), Wei et al. (2025) as well as Zakaria et al. (2024) demonstrate that blockchain-enabled supply chain integration positively affects green innovation efficiency, confirming the mediating role of integration. Similarly, Jie et al. (2024) develop a multi-factor explainable analytics framework showing that supply chain integration carries the effect of blockchain on green innovation. Zhao et al. (2023) further show that digital trust and task complexity moderate the blockchain–integration–green innovation pathway. These studies make important empirical contributions by establishing that mediation exists. However, they do not provide a theoretical explanation for why integration mediates, that is, what specific bundling functions integration performs and how those functions map onto the resource orchestration stages. The present study extends this line of research by grounding the mediation mechanism in resource orchestration theory’s three-stage process (structuring, bundling, leveraging), specifying three distinct systemic bundling functions of supply chain integration (complementary resource combination, systemic friction reduction, and dynamic responsiveness), and examining three boundary conditions (trust, task complexity, green digital learning orientation) as system-level contingencies that affect each orchestration stage. Thus, rather than merely confirming that integration mediates, this study explains how and under what systemic conditions it mediates.

2.4 Moderating variables

From a systemic perspective, the effectiveness of resource orchestration processes is contingent upon various system elements and contextual conditions that shape how resources are structured, bundled, and leveraged (Madlenak et al., 2025; Scott and Davis, 2015). This study examines three key moderating variables (supply chain trust, task complexity, and a green digital learning orientation) that represent different dimensions of the influence of the system environment on the relationship between BTA and green innovation.

2.4.1 Supply chain trust

Interorganizational trust refers to the positive expectations and confidence held by all parties involved in economic exchanges that they can effectively perform their duties and will not speculatively exploit the weaknesses of other parties (Gulati and Sytch, 2008). From a systemic perspective, trust functions as a relational governance mechanism that reduces systemic friction and enables more efficient resource flows across organizational boundaries (Poppo et al., 2016). Not only can it reduce uncertainty and opportunistic behavior in economic exchanges and lower transaction costs, but it can also effectively avoid moral hazard and promote interorganizational learning and joint problem solving, thereby improving organizational transaction performance and helping organizations gain competitive advantages (Gulati and Nickerson, 2008; Poppo et al., 2016). Supply chain partners who trust each other can create a more open and transparent trading environment (Li et al., 2021; Poppo et al., 2016), opening up all the data and information on the blockchain platform without worrying about the leakage of sensitive information and sharing only some of the data with partners. This trust-based openness enhances the system’s capacity for deep integration, as it reduces information asymmetry and enables more complete resource sharing across network nodes (Li et al., 2021). Consequently, supply chain trust accelerates the process through which enterprises apply blockchain technology to integrate their supply chains and improve the efficiency of resource bundling.

2.4.2 Task complexity

Applying a systemic lens, task complexity represents a structural characteristic of the system’s transactions that influences the codifiability and standardization of information flows. Task complexity describes the complexity of interorganizational transaction tasks and the degree to which the results of transaction tasks are difficult to standardize (Li et al., 2017; Luo et al., 2012). A transaction task requires a large amount of expertise and manpower from multiple fields, and a complex transaction task may contain a series of subtasks, making it difficult to measure the result of the transaction task by a unified set of standards (Daft and Macintosh, 1981; Mani et al., 2010). In the BTA context, overly complex tasks are difficult to convert into computer languages for coding (Daft and Macintosh, 1981; Mani et al., 2010). This conversion challenge introduces friction into the system, as the inherent requirement of the technology for codification conflicts with the inherent complexity and ambiguity of the task (Li et al., 2017). The increase in conversion time and cost hinders the process of BTA for supply chain integration and limits the efficiency of resource bundling, representing a systemic constraint on the digital transformation of interorganizational coordination.

2.4.3 Green digital learning orientation

From a systemic perspective, a green digital learning orientation represents a cultural-cognitive element of the organizational system that shapes the strategic priorities guiding resource allocation decisions. A green digital learning orientation is one way to use digital resources in an effective and eco-friendly manner (Benzidia et al., 2021), encouraging enterprises to incorporate energy-efficient practices and technology into the design and operation of servers (Costa and Opare, 2025) and other infrastructure components of digital learning platforms, with the aim of improving energy efficiency (Halbusi et al., 2024). Enterprises with this strong cultural orientation tend to have a long-term vision, uphold high ethical and moral standards and take responsibility for others and society (Wan et al., 2020). Consequently, such enterprises tend to avoid behaviors that do not meet ethical or moral standards, such as polluting the environment (Wan et al., 2020). Importantly, this orientation uniquely combines a focus on digital technology as the primary medium for learning, a strategic commitment to environmental sustainability, and an active learning process that transforms integrated data into green innovation outcomes (Zakaria et al., 2024). This orientation functions as a goal-setting mechanism that directs the system’s resource flows toward environmentally beneficial outcomes. Thus, it serves as the appropriate cultural-cognitive mechanism at the leveraging stage of resource orchestration, ensuring that resources bundled through supply chain integration are channeled toward sustainability objectives rather than being diverted to short-term cost reduction or other competing priorities. A strong corporate culture, that is, a green digital learning orientation, plays a pivotal role as a control mechanism that shapes enterprises’ behavior when they encounter ethical dilemmas and influences decision-making (Costa and Opare, 2025).

3 Hypotheses

3.1 The direct effect of BTA on green innovation

Enterprises operate as open systems, where technological innovation reshapes the patterns of resource acquisition and utilization across organizational boundaries (Madlenak et al., 2025). BTA functions as a digital infrastructure that fundamentally reconfigures how resources are identified, accessed, and structured within the supply chain network. This infrastructure generates emergent properties, such as system-wide resource visibility and tamper-proof transaction records, that cannot be reduced to any single firm’s internal capabilities. Within this systemic view, resource orchestration theory suggests that enterprises committed to coping with environmental changes through technological innovation are often able to enhance their resource development and utilization, thereby improving their innovation potential (Hakala, 2011).

Specifically, BTA plays an important role in the structuring phase of resource orchestration. The distributed ledger, smart contract, and immutable data traceability functions of blockchain technology enable the transparent identification and accurate allocation of resources across the supply chain system (Hastig and Sodhi, 2020). Unlike resource orchestration within a single firm, the systemic structuring of resources via BTA allows enterprises to access low-carbon materials, clean energy, and environmental technology data that are distributed across multiple partners, reducing search costs and validation uncertainties for green innovation. Moreover, as a decentralized database, blockchain technology fundamentally alters the interaction patterns within the supply chain system by breaking down barriers among participants (Sirmon et al., 2011). This systemic reconfiguration creates nonlinear feedback loops that increased data transparency encourages more partner participation, which in turn generates richer data for green innovation. With the advantages of trustlessness and traceability, blockchain technology is profoundly changing the operation mode of enterprises and green innovation practices. By optimizing business processes, reducing operating costs, and improving the efficiency of multiparty collaboration, blockchain technology has laid a solid foundation for creating a trusted system environment (Dong et al., 2025). Thus, through systemic resource structuring, BTA positively influences green innovation. Therefore, this study proposes the following hypothesis:

H1 BTA positively affects green innovation.

3.2 The mediating role of supply chain integration

The “structuring–bundling–leveraging” core framework of resource orchestration theory (Sirmon et al., 2007; Sirmon et al., 2011) provides a lens through which to understand how digital resources transform into advanced capabilities. Integrating this with a systemic perspective, supply chain integration represents a critical bundling mechanism that enables the transformation of structured BTA resources into green innovation through three interconnected pathways that operate simultaneously rather than in isolation.

First, through complementary resource combination, supply chain integration establishes cross-organizational collaboration networks that facilitate the interaction between blockchain data streams (such as supplier environmental certification and product carbon footprint labeling) and tacit knowledge (Ahmed et al., 2022; Liu et al., 2021). For example, blockchain-tracked waste recycling data can trigger suppliers and manufacturers to jointly develop new processes for recycled materials. Second, through systemic friction reduction, supply chain integration enables shared routines and coordinated workflows that lower monitoring costs and defensive behaviors, allowing fragmented blockchain resources to be pooled into complementary resource pools (Alyasein et al., 2025). Third, through dynamic responsiveness, supply chain integration addresses strategic rigidity by dynamically adjusting resource-bundling strategies in response to market or regulatory changes (e.g., policy adjustments increasing carbon taxes) (Iranmanesh et al., 2023; Liu et al., 2021). These three pathways are not mutually exclusive; rather, they co-evolve as the supply chain system gains experience with BTA. The relative dominance of each pathway depends on contextual factors such as trust and task complexity, but all contribute to the overall mediating effect. For example, when blockchain detects that a raw material supplier has ceased production because of an environmental incident, a highly integrated supply chain can quickly activate backup green suppliers and reoptimize production schedules, ensuring that the green innovation process is not interrupted. Thus, supply chain integration serves as a systemic bundling mechanism that fully transmits the effect of BTA to green innovation. Therefore, this study proposes the following hypothesis:

H2 Supply chain integration positively mediates the relationship between BTA and green innovation.

3.3 The moderating effect of supply chain trust

Supply chain trust functions as a relational governance mechanism that reduces systemic friction by lowering monitoring costs and defensive behaviors among members (Poppo et al., 2016). Trust does not simply substitute for blockchain’s technical assurance; rather, it amplifies the conversion of blockchain-enabled visibility into actionable integration by encouraging partners to disclose richer data, align decisions more quickly, and use shared blockchain records for collaborative problem solving rather than defensive monitoring. (Zhao et al., 2023). Supply chain trust also facilitates tacit knowledge exchange and encourages cross-boundary process alignment. When trust permeates the supply chain system, enterprises are more willing to open up full-link data obtained by blockchain technology to upstream and downstream partners, thereby supporting cross-organizational process restructuring, such as the joint design of low-carbon transportation routes (Dong et al., 2025). Consequently, by eliminating collaborative friction, supply chain trust strengthens the driving role of BTA in resource bundling, enabling deeper systemic integration. Therefore, this study proposes the following hypothesis:

H3 Supply chain trust positively moderates the relationship between BTA and supply chain innovation.

3.4 The moderating effect of task complexity

Situational constraints within the supply chain system significantly influence the effectiveness of resource bundling. Task complexity describes the complexity of interorganizational tasks and the difficulty of standardizing task results (Li et al., 2017; Luo et al., 2012). A complex task may contain a series of subtasks, making the results of the task difficult to measure with a uniform set of standards (Daft and Macintosh, 1981; Mani et al., 2010). A complex task requires a wide range of expertise in many fields, which may be difficult to encode through blockchain technology to upload to the database (Ziolkowski et al., 2020). This encoding difficulty impedes enterprises’ ability to structure data resources through BTA, creating systemic friction in the resource orchestration process. As task complexity rises, the marginal benefit of using blockchain for coordination declines because partners must devote more effort to managing mismatches between encoded logic and real-world task variability (Zhao et al., 2023). As a result, high task complexity, such as the need to coordinate multiple suppliers and dynamically adjust production plans for customized green products, leads to frequent supply chain interactions and increased uncertainty (Ziolkowski et al., 2020). From a systemic perspective, task complexity introduces structural constraints that limit the seamless integration of digital and human coordination mechanisms. Although BTA provides data transparency, the rigid contractual mechanisms of smart contracts do not respond quickly to dynamic changes in demand in complex tasks and may instead exacerbate process rigidity (Ante, 2020; Lumineau et al., 2021). For instance, BTA cannot renegotiate contract terms rapidly when environmental standards are temporarily adjusted. Thus, enterprises tend to rely on traditional flexible coordination methods, such as face-to-face communication, which weakens the enabling effect of BTA on supply chain integration. Therefore, this study proposes the following hypothesis:

H4 Task complexity negatively moderates the relationship between BTA and supply chain integration.

3.5 The moderating effect of a green digital learning orientation

Strategic guidance is needed to guide the release of resource value in the stage of capability leveraging (Chadwick et al., 2015). In the systemic framework, a green digital learning orientation serves as a goal-setting mechanism that directs resource flows toward sustainable innovation, ensuring that the leveraging process aligns with long-term environmental objectives (Scott and Davis, 2015; Halbusi et al., 2024). First, a green digital learning orientation couples digital learning processes with explicit environmental sustainability goals, creating a shared system-level norm that prioritizes green outcomes over operational efficiency alone (Costa and Opare, 2025; Halbusi et al., 2024). This strengthens the positive effect of supply chain integration on green innovation because integrated resources (such as supplier carbon data or customer eco-design feedback) are more likely to be channeled into environmental improvements rather than being diverted to short-term cost reduction (Wan et al., 2020). Second, it fosters cross-boundary experimentation and knowledge diffusion around low-carbon solutions, enabling the supply chain system to transform integrated data into collective green innovation capabilities (Zhou et al., 2025). For example, integrated blockchain-enabled information translates into joint R&D of cleaner processes or recyclable materials. Third, it institutionalizes learning routines that accelerate the conversion of bundled resources into tangible environmental improvements, including degradable material R&D and cleaner production processes, which ensures that the leveraging of integrated resources follows a systematic, repeatable pathway toward green innovation, rather than remaining ad hoc or efficiency-oriented. Thus, as the cultural-cognitive compass of the supply chain system, a green digital learning orientation positively moderates the effect of supply chain integration on green innovation. Hence, this study proposes the following hypothesis:

H5 A green digital learning orientation positively moderates the relationship between supply chain integration and green innovation.

4 Methods and analysis results

4.1 Sampling and data collection

The survey data were obtained from CEOs and top managers in the Chinese manufacturing sector and were especially concentrated in the metal, chemical and plastics, and mining industries. China provides a suitable environment for investigation because it is a large industrial country (Duan et al., 2021) and is keen on environmental protection (Guo et al., 2023).

Prior to data collection, unstructured interviews with five representative organizations were conducted, engaging their CEOs or other senior executives to understand the ideas of BTA and green innovation. Following these interviews and previous research, the English version of the questionnaire was initially created, subsequently translated into Chinese, and then retranslated into English by independent scholars to guarantee conceptual equivalence and precision, in accordance with the procedures recommended by Sekaran (1983). A pilot test was executed; it involved ten manufacturing companies that have used blockchain technology in their supply chains and solicited their CEOs to complete the questionnaire and assess its clarity and usefulness. Thus, the final version of the questionnaire was enhanced in accordance with their suggestions.

Owing to the challenges of coverage and participation in data collection within an emerging country, the Credamo platform was utilized to obtain survey data. This extensive online crowdsourcing platform, akin to the Qualtrics Online Sample, guarantees high-quality data by selecting participants based on demographics and excluding those with a history of providing low-quality responses (Zhou et al., 2023). The survey commenced in May 2025 and continued for 60 days. The questionnaires were distributed to the CEOs and senior executives of the companies.

Prior to distribution, we applied four screening criteria to ensure that only qualified respondents were invited: (a) the respondent’s firm must have implemented blockchain technology in its supply chain management; (b) the relationship with the firm’s top five suppliers must have lasted more than 1 year; (c) the respondent must be a CEO or senior manager with at least 3 years of tenure in the current position; and (d) the firm must have been using blockchain for supply chain management for a minimum of 12 months. These criteria were enforced by the Credamo platform’s targeting engine and by a filter question placed at the beginning of the questionnaire.

During the data collection phase, this study implemented measures to verify the informants’ qualifications and to mitigate social desirability bias. Initially, the informants were instructed to verify that their respective enterprises are business-to-business entities that have implemented blockchain technology in supply chain management. Second, the informants were requested to verify that the duration of the relationship between the questioned enterprise and its primary suppliers (top five) exceeded 1 year. Third, the questionnaires were stipulated to be completed by CEOs or senior management who were knowledgeable about both BTA and supply chain collaboration inside the organization, with all the respondents having more than 3 years of experience in their current roles. After two rounds of distribution, 452 completed questionnaires were obtained. The following sequential exclusion rules were applied: (a) removal of responses with missing values on any core construct item (55 cases); (b) detection of straight-lining (identical answers for more than 90% of Likert-scale items), this identified zero cases; (c) removal of responses that failed an embedded attention check question, Credamo’s initial screening had already removed 11 such cases, and we verified that no additional failures remained. After these exclusions, a final analytical sample of 397 valid responses was retained, yielding an effective response rate of 36.2% (397/1,250).

Non-response bias tests results are shown in Table 1. Early and late respondents (first 100 vs. last 100) on key firm characteristics and core variables were compared. Independent-samples t-tests showed no significant differences at the p < 0.05 level (e.g., for BTA: t = 0.83, p = 0.41). We also directly compared respondents (n = 397) with non-respondents (n = 853) on available firm-level attributes (size, industry, region). No statistically significant differences were found (e.g., industry distribution: χ² = 3.21, p = 0.52). Thus, non-response bias is unlikely to threaten the validity of the findings.

TABLE 1. Non-response bias tests.

Early respondents are defined as the first 100 returned questionnaires; late respondents are the last 100 returned questionnaires. Non-respondent data were obtained from the Credamo platform’s firm-level profile (size, age, industry, region).

Table 2 presents the sample information, encompassing enterprises of varying sizes, histories, and ownership types.

TABLE 2. Information of samples.

Common method bias was examined, with the results reported in Tables 3–6. To mitigate social desirability bias, this study guaranteed complete anonymity and employed more detailed and indirect questions throughout the survey (Podsakoff et al., 2003). Although procedural remedies were implemented during the survey design (e.g., anonymity, counterbalancing of question order, and clear instruction to reduce evaluation apprehension), this study conducted statistical tests to assess the potential influence of common method bias.

TABLE 3. Harman’s single-factor test.

The first factor explains 28.6% of the variance (<50% threshold), indicating no substantial common method bias.

TABLE 4. Marker variable test.

Marker variable: Attitude toward new technology (theoretically unrelated to the focal constructs). All adjusted correlations show minimal change, confirming that common method bias is not severe. **p <0.01.

TABLE 5. Variance inflation factor.

All VIF, values are below the conservative threshold of 3.3 (Kock, 2015), indicating no serious collinearity caused by common method bias.

TABLE 6. Common method factor approach.

Adding a latent common method factor does not significantly improve model fit, and the structural path coefficients remain virtually unchanged. All hypothesized relationships retain the same direction and significance levels. ***p < 0.001; n. s = not significant.

Harman’s single-factor test. All items of the focal constructs were entered into an unrotated exploratory factor analysis. The first extracted factor accounted for 28.6% of the total variance, which is well below the recommended threshold of 50%. This suggests that a single factor does not explain the majority of the covariance among the measures, indicating that common method bias is unlikely to be a serious concern.

Marker variable technique. Following the procedure recommended by Lindell and Whitney (2001), this study selected a theoretically unrelated marker variable (attitude toward new technology, Johnson et al., 2011; Richardson et al., 2009) that was not included in the structural model. This marker variable exhibited weak and non-significant correlations with all focal constructs (the highest absolute correlation was r = 0.08 with BTA, p > 0.05). After partialling out the marker variable’s influence, all significant correlations among the theoretical constructs remained significant and changed only marginally. This result further confirms that common method bias does not inflate the observed relationships.

Full collinearity assessment (VIF). Following the approach of Kock (2015), we computed the variance inflation factor (VIF) for each latent variable by regressing it against all other latent variables in the model. As shown in Table 10, all VIF values are below the conservative threshold of 3.3 (Kock, 2015), ranging from 1.84 to 2.91. This indicates that no substantial collinearity is present, which is consistent with the absence of serious common method bias.

Common method factor approach. We added an unmeasured latent common method factor to the measurement model, allowing all indicators to load on both their respective trait factors and the common method factor. The addition of the method factor did not significantly improve model fit (ΔCFI = 0.009, ΔRMSEA = 0.004). Moreover, the standardized path coefficients in the structural model changed by less than 0.05 after controlling for the method factor, and all hypothesized relationships remained significant in the same direction. This result suggests that common method bias does not materially alter the substantive conclusions of the study.

Taken together, the four complementary tests provide consistent evidence that common method bias is not a major threat to the validity of the findings.

4.2 Reliability and validity of the measures

Key variables were assessed via a seven-point Likert scale, with one indicating “strongly disagree” and seven indicating “strongly agree” (Table 7). The length of establishment (Ambulkar et al., 2015), annual revenue (Ambulkar et al., 2015), and trading frequency (Dong et al., 2019) were controlled for in the analysis.

TABLE 7. Measurement summary: Reliability and validity tests.

X²/df, Chi-square / degrees of freedom; RMSEA, Root mean square error of approximation; GFI, Goodness-of-fit index; CFI, Comparative fit index; NFI, Normed fit index; IFI, Incremental fit index.

Convergent validity. As shown in Table 7, all standardized factor loadings exceeded 0.70 and were significant at p < 0.001. Cronbach’s alpha for each construct exceeded 0.70, and composite reliability (CR) scores ranged from 0.761 to 0.842, well above the recommended threshold of 0.70. Average variance extracted (AVE) values ranged from 0.529 to 0.612, all exceeding the 0.50 benchmark (Fornell and Larcker, 1981). These results confirm adequate convergent validity for all constructs.

Discriminant validity. The Fornell-Larcker criterion was satisfied: the square root of each construct’s AVE was greater than its correlations with all other constructs. To provide a more rigorous test, we also employed the heterotrait-monotrait (HTMT) ratio of correlations (Henseler, et al., 2015). As reported in Table 8, all HTMT values are below the conservative threshold of 0.85 (the highest is 0.831 between SCI and GI), providing strong evidence of discriminant validity.

TABLE 8. HTMT discriminant validity assessment.

All HTMT, values are below 0.85, meeting the recommended threshold (Henseler et al., 2015).

Confirmatory factor analysis was conducted via AMOS 23.0. Table 7 presents the fit indices (χ2/df = 1.892; RMSEA = 0.034; GFI = 0.951; CFI = 0.917; NFI = 0.925; IFI = 0.938), demonstrating a good fit to the data. The structural model fit indices (SRMR = 0.058; d_ULS = 0.872) and the predictive relevance (Q2 = 0.312 for green innovation) are within acceptable ranges (Hair et al., 2019).

4.3 Analysis results

Table 9 presents the descriptive statistics and correlations of the key variables. All variables involved in the interactions were mean centered to mitigate any potential multicollinearity associated with testing the moderating effects.

TABLE 9. Descriptive statistics and correlations (N = 397).

*p < 0.05, p < 0.01, *p < 0.001.

The mediating and moderating effects were assessed via SmartPLS software (version 4.0) using bias-corrected bootstrap with 5,000 resamples. Table 10 illustrates the mediating effects, and the findings indicate that (1) BTA has a significant effect on green innovation (β = 0.146, p < 0.001), (2) BTA significantly affects the mediating variables of supply chain integration (β = 0.225, p < 0.001), and (3) when both the independent and the mediating variables are included in the regression model, the mediating variable of supply chain integration continues to significantly impact green performance (β = 0.177, p < 0.001), while the effect of BTA on green innovation becomes nonsignificant (β = 0.036, p > 0.05). As shown in Table 11, the indirect effect (BTA → SCI → GI) is β = 0.040, with a 95% bias-corrected confidence interval [0.019, 0.069], which does not include zero. The findings indicate that supply chain integration serves as a mediator of the relationship between BTA and green performance. Consequently, H1 and H2 are supported.

TABLE 10. The results of the mediating and moderating effects (N = 397).

*p < 0.05, p < 0.01, *p < 0.001.

TABLE 11. Bootstrapping results.

Table 10 also displays the findings regarding the moderating effects. Supply chain trust has a positive moderating effect on the relationship between BTA and supply chain integration (β = 0.342, p < 0.001), whereas task complexity has a negative moderating effect on the relationship between BTA and supply chain integration (β = −0.125, p < 0.01). At the same time, the moderating effect of a green digital learning orientation on the relationship between supply chain integration and green innovation is significantly positive (β = 0.219, p < 0.001). Consequently, H3, H4, and H5 are supported.

The following Table 11 presents the conditional indirect effect (mediated moderation) for each moderator, as well as the unconditional indirect effect of BTA on green innovation via supply chain integration.

5 Discussion and limitations

Grounded in resource orchestration theory and informed by a systemic perspective, this study provides empirical evidence of the fully positive mediating role of supply chain integration in the relationship between BTA and green innovation. It also demonstrates that supply chain trust strengthens the positive relationship between BTA and supply chain integration, whereas task complexity weakens it. In addition, green digital learning orientation positively moderates the relationship between supply chain integration and green innovation.

5.1 Theoretical implications

This study advances the literature by providing a system-level, processual account of how BTA enables green innovation through resource orchestration, theorizing the specific bundling functions of supply chain integration beyond mere mediation, and demonstrating that the BTA-green innovation link is contingent on system-level relational, structural, and cultural-cognitive factors.

First, this study integrates a systemic perspective with resource orchestration theory to specify how digital resources are transformed into system-level green innovation capabilities, moving beyond prior research that treated the BTA-green innovation link as direct or merely mediated. While previous studies have documented positive effects of BTA on green innovation (El-Kassar and Singh, 2019; Luo et al., 2023; Vaia et al., 2022), they have not explained the underlying transformation process from a resource orchestration standpoint. By explicitly mapping the three orchestration processes, that is, structuring (BTA-driven resource identification and allocation across the supply chain network), bundling (supply chain integration as a system-wide coordination mechanism), and leveraging (green innovation as an advanced capability), this study shows that green innovation emerges from the dynamic interactions among multiple actors within the supply chain system, not merely from a firm’s internal resource allocation. This processual, system-aware explanation contrasts with the linear input-output models that dominate prior literature and provides an analytical framework for understanding how digital technologies generate sustainable innovation through systemic resource orchestration.

Second, this study identifies supply chain integration as a bundling mechanism and, importantly, provides a theoretical explanation for why integration mediates the BTA-green innovation relationship. While Zakaria et al. (2024) and related empirical works (e.g., Jie et al., 2024) provide valuable evidence that blockchain-enabled supply chain integration positively affects green innovation efficiency, they focus on whether integration mediates rather than the underlying causal logic of how and why it does so from a theoretical standpoint. Drawing on resource orchestration theory, this study demonstrate that integration serves not merely as a channel for information transfer but as a systemic bundling process that performs three essential functions, namely, combining complementary resources (explicit, immutable data from blockchain with tacit, experience-based knowledge from supply chain partners), reducing systemic friction through shared routines and coordinated workflows, and enabling dynamic responsiveness to environmental changes (e.g., regulatory shifts, carbon price fluctuations). By theorizing these three bundling functions, our study refines the understanding of what supply chain integration actually does in the digital era, moving the literature from empirical verification of mediation to theoretical specification of bundling mechanisms.

Third, by examining three moderators that represent different dimensions of the system environment, namely, relational (supply chain trust), structural (task complexity), and cultural-cognitive (green digital learning orientation), this study reconciles prior contradictory findings about the effectiveness of BTA for green innovation and extends resource orchestration theory to incorporate system-level contingencies. Some studies have reported positive effects of BTA (Luo et al., 2023), while others have found negative or null effects (Najati, 2025; Zhan et al., 2025). Our contingency framework shows that BTA is not inherently beneficial or harmful; rather, its impact on supply chain integration (and hence on green innovation) depends on whether the system environment reduces friction (trust), avoids structural constraints (task complexity), and provides strategic direction (green digital learning orientation). This finding directly responds to the call for reconciling mixed results in the BTA literature. Moreover, it extends resource orchestration theory, which traditionally focuses on managerial actions within a single firm, by showing that the effectiveness of structuring (BTA), bundling (supply chain integration), and leveraging (green innovation) processes is contingent on system-level relational, structural, and cultural factors. This integration of system contingencies into resource orchestration theory opens new avenues for cross-organizational capability research.

5.2 Managerial implications

The empirical findings of this study present clear implications for managers in heavily polluting industries. The path from blockchain investment to green innovation passes through systemic supply chain integration, supported by trust, moderated by task complexity, and guided by a green digital learning orientation. Implementing blockchain as an isolated technical solution is insufficient; rather, it must be embedded within a broader organizational and interorganizational system that aligns technological, relational, and cultural elements toward sustainability goals.

First, managers should leverage blockchain not merely as a record-keeping tool but as a systemic integration infrastructure that enables real-time, tamper-proof environmental information sharing across the supply chain. Our finding that supply chain integration fully mediates the BTA-green innovation link implies that the value of blockchain lies not in isolated adoption but in its capacity to structure and bundle distributed resources. Concretely, manufacturing enterprises can deploy blockchain platforms to record immutable data on raw material origins, production process emissions, and product carbon footprints, then share this data with upstream suppliers and downstream partners. Unlike conventional databases, blockchain’s immutability builds a shared factual baseline that reduces asymmetric information and opportunistic behavior, enabling collaborative green innovation. For example, by tracking the carbon intensity of each supplier’s materials on a blockchain, a manufacturer can identify the most sustainable sources and work with suppliers to redesign production processes for lower overall emissions, which is a form of systemic resource bundling that is difficult to achieve with traditional information systems.

Second, managers should exploit smart contract functionality to automate routine green compliance tasks and reduce transaction friction. Smart contracts can encode environmental standards (e.g., emission thresholds, recycling content requirements) and automatically execute actions when conditions are met, such as releasing payment only after a supplier provides verifiable green certification or triggering penalties for non-compliance. This automation reduces costly manual monitoring and, importantly, builds supply chain trust by making commitments self-enforcing. Our finding that supply chain trust positively moderates the BTA-supply chain integration relationship suggests that trust and blockchain are complements rather than substitutes. By automating verifiable green tasks (e.g., verifying that recycled materials meet quality standards, that carbon offsets are properly retired), smart contracts free managerial attention for more complex, non-routine green innovation activities that require human judgment and negotiation.

Third, managers must recognize the limits of blockchain under high task complexity. Our finding that task complexity weakens the positive effect of BTA on supply chain integration implies that blockchain is not a panacea. When green innovation involves highly customized products, frequent engineering changes, or multi-party negotiations that cannot be easily encoded into computer code, over-reliance on blockchain may create process rigidity and reduce efficiency. In such situations, managers should adopt a hybrid governance approach. For instance, they should use blockchain for standardized, codifiable parts of the supply chain (e.g., tracking baseline emissions, certifying standard materials), while supplementing with traditional flexible coordination mechanisms for complex, ambiguous tasks (e.g., regular cross-organizational meetings, relationship managers, joint problem-solving sessions). This balanced strategy ensures that blockchain enables rather than hinders green innovation.

Fourth, trust building is a prerequisite for realizing the full value of BTA. The positive moderating effect of supply chain trust indicates that blockchain alone cannot force deep collaboration if trust is absent. Managers should invest in trust-building activities—such as joint problem-solving sessions, transparent sharing of non-competitive environmental data, and long-term relationship commitments—before or in parallel with blockchain implementation. When trust is present, partners are willing to share sensitive environmental data (e.g., energy consumption, waste generation, proprietary process information) on the blockchain, enabling the deep resource sharing that drives systemic green innovation. Conversely, in low-trust environments, even the most sophisticated blockchain system will yield only superficial integration.

Finally, managers should cultivate a green digital learning orientation as a strategic compass that directs the leveraged resources toward environmental outcomes. Our results show that this orientation positively moderates the supply chain integration-green innovation relationship, meaning that integrated resources will not automatically translate into green innovation without a clear sustainability-focused cultural-cognitive framework. Managers can embed this orientation through concrete mechanisms, such as setting shared green innovation KPIs for all supply chain partners, organizing regular workshops where blockchain-enabled environmental performance data are collectively discussed to derive improvement actions, and rewarding partners that actively share green digital lessons learned. These practices ensure that the resources bundled via blockchain and supply chain integration are channeled toward green innovation rather than diverted to short-term cost reduction or other competing priorities.

5.3 Limitations and future research

Despite its contributions, this study has several limitations that should be acknowledged and addressed in future research.

First, the data of this study were collected at a single point in time, so the hypothesized directions (BTA → supply chain integration → green innovation) are based on theoretical reasoning rather than longitudinal evidence. Although this study controlled for several confounding variables and tested mediation using bootstrapping, reverse causality or omitted variable bias cannot be entirely ruled out. Future research should use longitudinal or quasi-experimental designs (e.g., panel data, difference-in-differences) to establish causal ordering.

Second, common method bias may inflate the observed relationships. Although this study employed procedural remedies and statistical tests indicated that common method variance was not severe, future studies should collect data from multiple informants or triangulate with objective blockchain transaction data.

Third, the measurement of blockchain technology application is limited. Three-item scale assesses the degree to which firms use blockchain for key activities—specifically, operations, coordination, and transaction settlement. Although, in practice, the blockchain networks established among enterprises are typically consortium blockchains, future research can attempt to distinguish between blockchain types (public, private, consortium), the depth of integration with existing enterprise systems (ERP, SCM), or deployment maturity (Zulkifli et al., 2025), and develop more fine-grained measurement methods to capture the heterogeneity in blockchain adoption. In addition, future research could also explore the integration of blockchain with other cutting-edge technologies, such as large language models in supply chain finance (Yang et al., 2025), to examine how multi-technology synergies affect green innovation efficiency.

Fourth, supply chain integration was measured only from the supplier side, ignoring customer integration and internal integration. This study focused on upstream integration because the theoretical mechanism emphasizes resource bundling from suppliers (raw materials, environmental data, green technologies). However, in many industries, customer integration (e.g., sharing green product requirements, co-designing sustainable packaging) and internal integration (e.g., cross-functional collaboration within the firm) also matter for green innovation. Future research should adopt a multi-dimensional measure of SCI.

Fifth, the sample is restricted to heavily polluting industries in China. The findings may not generalize to low-pollution sectors, service industries, or other institutional contexts (e.g., less regulated markets, different cultural settings). Future studies should test the model in diverse industry and national contexts.

Finally, potential endogeneity concerns remain. Although this study included several control variables, unobserved factors could simultaneously drive BTA adoption, supply chain integration, and green innovation. Future research should address endogeneity more rigorously, for instance, by using policy shocks or blockchain infrastructure rollouts as natural experiments.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Pei Xiao, University of Surrey, United Kingdom

Reviewed by: Lei Yang, Shenyang University of Technology, China

Jiacheng Dong, Instituto Tecnologico De Chiná, Mexico