The Impact of Supply Chain Structure Diversification on High-Quality Development: A Moderating Perspective of Digital Supply Chains

Abstract

This study examines how supply chain structure diversification drives high-quality enterprise development in the digital economy. Using panel data from Chinese listed non-financial firms (2009–2023), we find that diversification of both suppliers and customers significantly improves firms’ total factor productivity (TFP), and the results remain robust after controlling for endogeneity. Mechanism analyses show that diversification enhances innovation capability, sustainability performance, and risk resilience, while digital supply chains strengthen these effects by improving information flow and coordination. Heterogeneity tests reveal that the impact is greater for firms with higher operational efficiency, cultural synergy, and information transparency. Overall, the findings highlight that diversified and digitally integrated supply chains are essential for innovation-driven, resilient, and sustainable enterprise growth.

1. Introduction

In the context of global climate change, shifting geopolitical dynamics, advancements in digital technologies, and the repercussions of public health events, traditional supply chain centralization strategies increasingly fall short of meeting the sustainable development requirements of enterprises. Consequently, supply chain structural diversification has evolved from being viewed as an “optional strategy” to a “necessity for survival” [1]. Under the pressures of persistent systemic risks, accelerating technological transformation, green development imperatives, and increasingly personalized market demands, enterprises must shift from an “efficiency-first” model to one that balances resilience and innovation. This transformation is achieved through optimized resource allocation, expanded capability boundaries, and extended value chains [2]. For example, in the early 2010s, Apple Inc. operated a highly centralized supply chain concentrated in mainland China. To mitigate the risks of over-concentration, Apple has since adopted a global “China + N” distributed model, characterized by “in-house technology development + regional diversification + ecosystem collaboration.” This strategic shift highlights the crucial role of supply chain restructuring in enhancing firms’ risk resilience and technological autonomy [3,4]. As a result, supply chain structure diversification has increasingly become a key pathway for enterprises worldwide to achieve high-quality economic development [5].

High-quality development represents a fundamental transformation in the stage of economic growth—from input-driven expansion to productivity-driven advancement [6]. Total Factor Productivity (TFP) measures the growth in output that results from non-factor inputs such as technological progress, managerial optimization, and improved resource allocation, beyond the contributions of traditional inputs like capital and labor. TFP thus reflects the efficiency and quality of growth [7]. It enables enterprises to achieve sustainable expansion that is not limited by factor constraints, directly addressing the core principle of efficiency in high-quality development [8]. The growth trajectory of listed companies typically follows two paths: one driven by factor accumulation and another by total factor productivity. The former relies on increased tangible inputs—capital and labor—leading to diminishing returns, overcapacity, and inefficiency, which conflict with the objectives of high-quality development [9,10,11]. The latter depends on intangible improvements such as innovation, management optimization, and effective resource reallocation. This productivity-driven approach enhances efficiency and supports sustainable growth [12,13,14]. Therefore, TFP levels and trends serve as direct indicators of whether enterprises have transitioned from factor-dependent to efficiency-driven growth, making TFP a core metric of high-quality development [15].

The high-quality development of listed firms goes beyond mere scale expansion or short-term profitability. It reflects a deeper transformation that prioritizes innovation, efficiency, sustainability, value creation, and effective risk management [16]. Because TFP quantifies the contribution of non-factor inputs, it directly aligns with the essence of high-quality development [17]. We posit that improvements in TFP depend primarily on three interrelated capabilities: innovation, sustainability, and risk resilience. First, innovation capability enables breakthroughs that raise the efficiency ceiling and determine the extent of productivity gains [18]. Second, sustainability capability provides long-term support by ensuring the durability of those efficiency gains [19,20]. Third, risk resilience underpins the stability of productivity by safeguarding firms against shocks [21]. Cultivating these three capabilities is therefore essential for achieving high-quality enterprise development.

Supply chain structure diversification refers to the strategic design and management of supply chains through the inclusion of multiple suppliers, logistics channels, production bases, and sales networks, forming a multi-dimensional, multi-layered system. This structure enhances supply chain resilience, risk resistance, and market adaptability [22]. The core principle is to reduce dependence on single entities by decentralizing supply chain layouts, thereby lowering potential risks while integrating diverse resources to optimize efficiency [23]. The resulting value extends throughout the supply chain and benefits all participants [24]. From short-term risk mitigation to long-term capability enhancement, diversification collectively steers the supply chain ecosystem toward high-quality development. For partner enterprises, the key is to leverage “collaborative empowerment” to transform external support into internal capability rather than relying solely on orders from leading firms [25]. Thus, supply chain diversification represents an ecological win–win strategy rather than an exclusive advantage for core firms. It fosters innovation by providing an external knowledge pool, promotes sustainability by building greener and more adaptive networks [26], and strengthens risk resilience through flexible system design. Enterprises should regard diversification as a strategic investment aimed at achieving a “1 + 1 + 1 > 3” synergy through capability integration [27].

Moreover, given the influence of large firms on their suppliers and customers—especially small and medium-sized enterprises (SMEs)—the diversification strategies adopted by leading companies can transmit risks and opportunities throughout the supply chain. Our analysis shows that digital transformation amplifies these spillover effects by enhancing information sharing and ecosystem efficiency. Prior studies demonstrate that customer adoption of artificial intelligence improves suppliers’ capacity utilization through more effective information flows [28], and that the integration of social responsibility and green principles produces strong positive spillovers for suppliers [29]. Li et al. (2025) and Liang et al. (2025) further show that digital transformation in leading firms enhances information transmission and resource allocation efficiency across the entire supply chain [30,31]. Greater diversification also reduces single-dependency risks, as digital information networks and precise resource-matching mechanisms improve collective resilience. Overall, digitalization lowers the likelihood that core firms shift risks onto partners, thereby enhancing systemic stability—highlighting the importance of digital supply chains as a moderating factor.

In light of this context, this study explores whether and how supply chain structure diversification promotes high-quality development among micro-enterprises in the digital economy. Using econometric models and empirical analysis based on Chinese listed non-financial firms from 2009 to 2023, we examine the direct impact of supply chain diversification on high-quality development and the mechanisms through which it operates. Our findings show that supply chain diversification significantly promotes high-quality development by improving firms’ innovation capability, sustainability performance, and risk resilience. We further find that supply chain digitalization amplifies this positive effect. The relationship is stronger among firms characterized by high operational efficiency, cultural synergy, and information transparency.

This study makes three main contributions. First, it examines the impact of supply chain structure diversification on high-quality development from a structural perspective, enriching the literature on the determinants of enterprise productivity and sustainability. By identifying the channels of innovation, sustainability, and risk resilience, it provides new insights into how supply chain configurations foster technological progress and efficiency optimization. Second, it extends the research on the economic consequences of supply chain diversification to include firm-level high-quality development, considering both upstream and downstream perspectives. Third, it identifies moderating mechanisms—digitalization, efficiency, culture, and information symmetry—that magnify the benefits of diversification. Together, these findings offer novel strategies for improving supply chain stability and promoting sustainable, high-quality economic growth.

The remainder of this paper is organized as follows. Section 2 reviews the relevant literature and develops the research hypotheses. Section 3 describes the data and model specification. Section 4 presents the empirical results and analysis. Section 5 concludes with key findings and implications.

2. Literature Review and Hypothesis

2.1. Supply Chain Diversification and High-Quality Development

As global supply chains continue to evolve toward diversification, regionalization, and digitalization, fostering innovation capability, sustainability, and risk resilience at the enterprise level—microeconomic actors crucial to economic and societal development—has become a critical pathway for enhancing total factor productivity (TFP) and advancing high-quality development. Existing studies on the economic effects of supply chain diversification suggest that it helps firms reduce capital costs and inefficient investments [32,33], improve sustainability performance [26], enhance innovation outcomes and green innovation [34,35], and strengthen risk-taking and digital transformation [36,37]. Regarding factors influencing high-quality development, prior literature has highlighted the roles of the digital economy [38,39], social capital [40], financial innovation [41], green governance [42], green innovation capability [43], supply chain ownership [44], artificial intelligence [45], digital finance [46,47], education levels [48], green financial innovation [49], and cultural heritage preservation [50], all of which contribute to advancing high-quality development. However, the following question arises: how does supply chain diversification improve TFP and thereby facilitate high-quality development?

Drawing upon resource dependence theory, enterprises can reduce reliance on singular supply and sales channels through supply chain diversification, thereby mitigating the adverse impacts of external shocks on production processes [1]. This decentralization strategy lowers the likelihood of supply chain disruptions, ensuring production continuity and minimizing capacity idling caused by resource shortages. From the perspective of production functions, stable factor supplies enable actual utilization rates of capital and labor to approach their theoretical optimal values [51]. Moreover, the integration of heterogeneous resources fosters unique, hard-to-imitate competitive advantages, accelerating the growth rate of technological advancement in the production function [52]. Consequently, supply chain diversification enhances total factor productivity and facilitates high-quality development. According to transaction cost economics, enterprises that establish “competitive supply chain networks” through multi-sourced procurement and multi-channel sales can simultaneously achieve the price advantages of scale purchasing and the flexibility benefits of adaptive adjustments. This diversified structure prevents monopolistic pricing by single suppliers or customers and avoids losses in economies of scale, while maintaining overall purchasing and selling volumes through dynamic order allocations. By reducing marginal input costs and improving factor combination efficiency on the output side, this formation effectively enhances total factor productivity, thereby advancing high-quality development [53]. Within the framework of information economics, diversified supply chains furnish enterprises with multidimensional market signal inputs. Price fluctuations and delivery cycle variations from different suppliers or customers create a cross-validated information network that enables core enterprises to predict trends in factor markets with greater accuracy [54]. This informational advantage allows enterprises to minimize trial-and-error costs associated with production planning and technological pathways [55], thereby enhancing the foresight of factor allocation. This improvement is directly reflected in increased management efficiency within total factor productivity, subsequently promoting high-quality development in enterprises [56].

Ultimately, supply chain structural diversification systematically optimizes factor allocation efficiency and accelerates the rate of technological advancement through mechanisms such as risk diversification, reconstruction of economies of scale, and information optimization [57]. This process not only enhances total factor productivity but also effectively promotes high-quality economic development [58,59]. Furthermore, this approach aligns with the principles of total factor productivity outlined in neoclassical growth theory and resonates with the adaptive improvements emphasized by evolutionary economics. Accordingly, this paper proposes the following research hypothesis:

H1. 

Ceteris paribus, supply chain structural diversification can significantly promote high-quality development in enterprises.

2.2. The Mediating Role of Innovation Capability

Collaborative innovation acts as an “amplifier” of innovation capability; it is not only a key means for enterprises to overcome internal resource constraints but also directly reflects their ability to acquire, integrate, and transform innovative elements within an open ecosystem. Knowledge diversification serves as the “underlying code” of innovation capability, essentially evolving organizations into innovative ecosystems that can continuously absorb, integrate, and create diverse knowledge.

Regarding the promoting effect of supply chain structural diversification on collaborative innovation and knowledge diversification, this study posits that core enterprises enhance innovation efficiency through resource coupling, thereby facilitating the expansion of knowledge networks. Simultaneously, they strengthen learning capabilities through network diffusion, driving the evolution of knowledge structures and ultimately establishing a spiraling pathway of innovation output and knowledge accumulation. This dynamic adaptation mechanism enables enterprises to transcend the linear innovation limitations of traditional supply chains, achieving reciprocal empowerment of collaborative innovation and knowledge evolution. The specifics are as follows: First, the heterogeneous resource coupling mechanism. The multidimensional expansion of supply chain nodes forms a resource network that spans industries and technological domains, leading to a coupling effect of diverse resources with different knowledge schemas at organizational boundaries [60]. Based on the standardization of technological interfaces, enterprises integrate discrete specialized capabilities into innovative combinatorial advantages through the modular decomposition and reconstruction of knowledge units. This resource coupling breaks the path dependence of a single knowledge trajectory, shifting the direction of innovation from linear extension to multidimensional leap [61]. Second, the networked information diffusion mechanism. The multi-centered network topology created by diversified supply chains generates super-linear growth information channels. The presence of structural holes allows enterprises to occupy intermediary positions, enabling them to simultaneously access diverse, asymmetric knowledge flows from multiple fields. This cross-boundary information transfer generates a knowledge recombination effect, where tacit experiences are encoded into transferable innovative elements through social networks, catalyzing knowledge integration across adjacent technological domains and thus forming an exponentially amplified space for innovation potential [62].

Regarding the driving effects of collaborative innovation and knowledge diversification on enterprise total factor productivity (TFP), this study suggests that the former provides incremental exogenous knowledge, while the latter ensures structural flexibility of endogenous knowledge. Specifically: First, from the perspective of collaborative innovation, the significance of technological integration and risk diversification. Collaborative innovation breaks organizational boundary barriers, enabling cross-entity knowledge sharing that facilitates modular technological complementarity. In the process of innovation synergy, differentiated technological accumulation among heterogeneous entities generates a superadditive effect, surpassing the threshold limitations of individual entities’ R&D capabilities [63] and reducing the marginal costs of technological exploration [64]. Furthermore, the risk-sharing mechanisms among multiple entities weaken the negative impacts of technological uncertainty, allowing enterprises to sustain high R&D intensity at lower costs, thereby providing technological support for high-quality development [65]. Second, from the perspective of knowledge diversification, cognitive reconstruction and efficiency improvement. The cross-penetration of diversified knowledge elements breaks existing cognitive stereotypes, enabling enterprises to generate emergent innovations through conceptual reorganization. The nonlinear interactions among heterogeneous knowledge units lead to a supermodel effect, where the outward expansion of the production possibility frontier exceeds the cumulative results of single-dimensional knowledge additions, thereby facilitating the scaled growth of enterprise TFP. Moreover, the reserve of diversified knowledge creates a buffering mechanism against market disturbances, shortening the lag time for strategic adjustments, enhancing the reallocation efficiency of production factors, and improving the economic quality of enterprise development.

Based on this analysis, the second research hypothesis is proposed:

H2. 

Under ceteris paribus conditions, supply chain structural diversification significantly enhances innovation capability, thereby promoting high-quality enterprise development.

2.3. The Mediating Role of Sustainable Development Capability

ESG performance serves as a fundamental framework for evaluating a company’s sustainability capabilities, assessing its potential for long-term value creation in non-financial domains, and revealing its capacity for achieving stable growth within the confines of environmental constraints, social demands, and governance norms [66]. Supply chain structural diversification can systematically enhance a company’s overall performance across environmental, social, and governance dimensions. This is accomplished by improving environmental performance through resource integration, bolstering social resilience via risk diversification, and optimizing governance structures through knowledge spillovers [67]. The facilitative effect of this approach can be primarily observed through three core mechanisms. First, the integration of heterogeneous resources and green synergy effects. Supply chain diversification introduces partners from various industries and technological fields, integrating resources with different environmental attributes into the supply chain network. This diversified structure reduces dependence on single high-pollution links and prompts upstream and downstream enterprises to collaboratively establish green standards and promote sustainable manufacturing practices, thereby forming a closed loop of environmental responsibility throughout the product lifecycle [68]. Second, risk diversification and enhanced resilience. A diversified supply chain diminishes the risks of operational disruptions stemming from localized environmental disasters or social conflicts by broadening geographic distribution and diversifying supply chain partner types, thus maintaining stable capacity for fulfilling social responsibilities. Third, knowledge spillovers and enhanced governance capacity. A diversified supply chain network facilitates the exchange of knowledge paradigms from various fields (e.g., green technologies, social responsibility management), creating a distributed learning mechanism. This enables enterprises to translate environmental protection technologies and social experiences into reusable solutions within the supply chain [69]. This knowledge-sharing mechanism not only enhances environmental performance but also strengthens corporate governance capabilities through collaborative innovation [70].

ESG performance reconstructs the combinatorial efficiency of production factors and growth models through three pathways: technological innovation, resource allocation, and risk management, ultimately driving a paradigm shift in total factor productivity from scale-driven to quality-driven [71]. First, the impact on driving technological innovation. ESG responsibilities motivate enterprises to invest in technological innovation, transcending the technical boundaries of traditional production functions [72]. Environmental obligations compel enterprises to develop low-carbon technologies [73], social responsibilities drive innovations in products and services that cater to societal needs [74], and governance responsibilities ensure efficient R&D resource allocation through optimized decision-making processes. Furthermore, ESG-driven green technological innovation exhibits superadditive characteristics, reducing both marginal costs and environmental externalities, thereby fostering a positive feedback loop for technological advancements [75]. Second, the optimization of resource allocation efficiency. ESG information disclosure mitigates information asymmetry and decreases inefficient investment behaviors, thereby enhancing capital allocation precision [76]. The fulfillment of social responsibilities boosts employees’ sense of belonging and creativity [77], while optimized governance structures improve the synergy efficiency of human capital, resulting in increased labor productivity [78]. Green supply chain management encourages upstream and downstream enterprises to upgrade technologies, thereby enhancing factor utilization efficiency across the entire chain through industry network externalities [79,80]. Third, risk hedging and cost restructuring. ESG practices lead to reductions in non-productive losses by lowering environmental regulation risks, social conflict costs, and the likelihood of governance failures [81]. Enhanced environmental performance reduces compliance costs such as carbon taxes and pollution fees, social performance minimizes production interruptions caused by labor disputes, and governance performance curtails the erosion of capital returns due to agency costs. This risk-hedging effect liberates additional resources for productive investment while lowering capital costs through green financing channels, thus creating a virtuous cycle of “cost savings–efficiency enhancement–value creation” [82]. In light of these insights, the following research hypothesis is proposed:

H3. 

Under ceteris paribus conditions, supply chain structural diversification significantly strengthens sustainability capabilities, thereby promoting high-quality enterprise development.

2.4. The Mediating Effect of Risk Resilience

Profit volatility serves as a critical indicator of a company’s risk resilience, reflecting the dynamic equilibrium between risk absorption capacity and risk exposure. Low volatility suggests that a company possesses a robust structure capable of withstanding shocks and demonstrating strong risk resilience.

Supply chain structural diversification systematically reduces profit volatility by establishing a multidimensional and multilayered networked resource allocation system [83]. The mechanisms underlying this reduction can be decomposed into three interrelated levels. First, the mechanisms of risk diversification and shock buffering. By incorporating a network of suppliers and customers from various regions and industries, companies can create redundant supply and sales pathways. When a specific node encounters external shocks, other nodes can seamlessly take over, ensuring production continuity and preventing significant income declines associated with disruptions in traditional single supply chains. Furthermore, cyclical variances across regional markets and the technological characteristics inherent in various industries provide natural hedges that mitigate the impact of market fluctuations on overall profits, thereby achieving a dynamic balance in income structure [84]. Second, the integration mechanisms for resource heterogeneity. By uniting suppliers and customers with diverse technological paths and resource endowments, companies can construct a more resilient resource pool capable of withstanding recessions. When the value of a particular resource type is diminished by market fluctuations, value can be extracted from other resources to compensate for income losses. Additionally, the competitive bargaining environment generated through multi-sourced procurement and sales diminishes the price monopoly risks associated with single suppliers or customers, thereby reducing the rigidity of fixed costs that can otherwise negatively impact profits [85]. Third, the mechanisms for enhancing information synergy and responsiveness. The multi-node information touchpoints of a diversified supply chain network can concurrently detect early signals of fluctuations across different markets and technological domains. By cross-validating big data, companies can more accurately anticipate trends, allowing for timely adjustments to production plans and inventory strategies [86]. Moreover, the networked flow of information diminishes the time-lag effects of traditional linear decision-making processes. Utilizing intelligent algorithms for real-time analysis of multi-source data enables companies to swiftly reconstruct and alter supply chain pathways, thereby controlling the impact of external shocks on profits within the critical decision-making time window.

Profit volatility enhances a company’s total factor productivity through three mechanisms: precise innovation activation, rational decision reinforcement, and resource efficiency reconstruction. This fundamentally reflects the transformation of market uncertainty into organizational adaptability value. First, the mechanisms of stress response and innovation activation. Profit volatility acts as an amplifier of market signals, compelling companies to transcend technological path dependency. The survival pressures induced by volatility initiate an “innovation compensation effect,” leading to increased R&D investments in high-risk, high-reward areas. By broadening their technological search space, companies can escape local optimum traps and achieve Pareto improvements in their technological portfolios [87]. Second, the reinforcement of decision rationality. Continuous volatility prompts decision-makers to revise their prior probability distributions, enhancing cognitive accuracy through improved information discrimination. Furthermore, volatility enhances companies’ management of real options, transforming uncertainty into strategic flexibility via phased investments and the modular development of technology. This decision-making model facilitates more efficient dynamic adjustments in R&D projects, thereby mitigating the erosion of total factor productivity caused by the sunk cost effect. Third, the reconstruction of resource efficiency. Profit volatility, as a proxy for market selection mechanisms, accelerates the identification and divestment of inefficient assets. By triggering creative destruction processes, companies can optimize the ratio of asset specificity to generality, thereby enhancing the adaptive efficiency of factor combinations. Additionally, volatile environments strengthen the resilience of talent teams in challenging situations, filtering out highly adaptable individuals through self-selection mechanisms, which subsequently translates human capital elasticity into productivity gains [88]. Based on these insights, the following hypothesis is proposed:

H4. 

Under ceteris paribus conditions, supply chain structural diversification significantly enhances risk resilience, thereby promoting high-quality enterprise development.

2.5. Moderating Effect of Supply Chain Digitalization

Supply chain digitalization has emerged as a critical driver of superior supply chain performance [89]. According to the resource orchestration theory, as a higher-order dynamic capability, digital technologies amplify the value of structural diversification by reconfiguring resource orchestration [90]. Digital elements complement traditional supply chain components, generating synergistic effects [91]. Additionally, algorithm-driven intelligent decision-making systems enhance the dynamic adaptability of resource orchestration, while platform-based architectures create new value generation pathways. This technology–organization co-evolution mechanism positions supply chain digitalization as a pivotal lever for leveraging structural diversification to improve total factor productivity [92,93]. Specifically, first, structural optimization of the information synergy mechanism. The network complexity introduced by supply chain diversification increases information transmission layers and coordination costs among nodes. By adopting digital technologies, firms build data hubs and intelligent algorithm platforms, enabling standardization, integration, and real-time sharing of multi-source heterogeneous data, thereby eliminating the “information silos” inherent in traditional supply chains [94]. The immutability of blockchain technology enhances the traceability of transactions across entities, while real-time operational data collected by IoT devices improves supply chain transparency [95]. This upgraded information synergy mechanism significantly reduces transaction friction costs in diversified structures, allowing firms to more efficiently allocate dispersed resources within complex networks [96]. Second, a paradigm shift in dynamic decision-making capability. When diversified supply chains encounter uncertainty risks, such as demand fluctuations or supply disruptions, traditional experience-based decision-making models often suffer from delayed responses. Firms employing digital systems leverage machine learning algorithms to conduct in-depth analysis of historical data, develop multidimensional predictive models, and proactively identify potential risk points [97]. Virtual simulation systems powered by digital twin technology enable stress testing and pathway optimization for diversified supply chain networks. By integrating predictive decision-making capabilities with real-time dynamic optimization mechanisms, firms transcend the reactive limitations of traditional supply chain management, achieving proactive resource allocation efficiency in diversified structures [98]. Third, a marginal breakthrough in knowledge integration efficiency. The implicit knowledge embedded in diversified supply chain networks is often decentralized and heterogeneous. Natural language processing technologies can automatically extract specialized knowledge from unstructured documents, knowledge graph technologies construct cross-domain conceptual association networks, and deep learning models facilitate the transformation of tacit knowledge into explicit knowledge. The establishment of this digital knowledge management system overcomes the bottlenecks of knowledge transfer constrained by interpersonal communication in traditional supply chains, enabling firms to systematically integrate technical expertise, management experience, and market insights from diverse networks, thereby converting them into innovative momentum for continuous production process improvement [99]. Fourth, institutional restructuring for ecological collaborative innovation. Digital platforms disrupt the linear relationships of traditional supply chains by creating a networked collaboration ecosystem involving multiple stakeholders [100]. Smart contract technologies lower collaboration costs through the automated execution of predefined rules, while cloud-based collaborative design platforms facilitate the sharing of R&D resources across organizations [101]. This digital-driven open innovation model allows firms to transform customers, suppliers, R&D institutions, and other entities in diversified supply chains into innovation communities. Through continuous technological iterations and process improvements, firms achieve Pareto optimization of total factor productivity [102]. Based on this, the fifth research hypothesis is proposed:

H5. 

In the process of supply chain diversification driving high-quality enterprise development, supply chain digitalization exerts a significant positive moderating effect.

3. Data Selection and Model Setting

3.1. Sample Selection and Data Sources

This study selects listed companies in China’s non-financial sector from 2009 to 2023 as the initial research sample for several reasons. First, the changing policy environment. Following the global financial crisis in 2008, governments worldwide implemented a series of economic stimulus policies and reforms, which profoundly influenced corporate operations and supply chain management. Second, the economic cycle phase. The data from after 2009 encompasses various stages of the capacity cycle, allowing for a comprehensive reflection of the role of supply chain structural diversification in promoting high-quality enterprise development across different economic contexts. Third, data availability and completeness. Over time, the mechanisms for collecting and organizing corporate data have improved, making data from 2009 onwards more accessible and complete. Finally, the acceleration of digital transformation in enterprises. After 2009, the rapid advancement of information technology accelerated the digital transformation process of companies, significantly increasing data generation and accumulation. This enables companies to more accurately record and reflect their supply chain structures and operations, providing researchers with richer and more accurate data to measure the relevant indicators of supply chain structural diversification and high-quality enterprise development. The data sources for the sample companies are divided into three categories: (1) Basic information, financial indicators, and corporate governance data from the CSMAR database. (2) Cooperative innovation patent data from the Chinese National Research Data Service Platform (CNRDS), patent data on knowledge diversification from the National Key Industries Patent Information Service Platform, ESG data from the Shanghai Huazheng Index sourced from the Wind database, and internal control indices from the Dibao database. (3) Textual data on supply chain digitization, cooperative culture, and supply chain finance from the annual reports of listed companies. Additionally, the initial research sample underwent the following treatments: (1) Exclusion of companies that were delisted within the study period; (2) Exclusion of companies marked with abnormal statuses such as ST or *ST in the same year; (3) Exclusion of samples with undisclosed or incomplete financial data. After sorting, the study collected a total of 36,344 company-year observations. To mitigate the influence of outliers, winsorization was applied to all continuous variables at the 1% level.

Regarding sample selection, this study focuses on Chinese A-share listed companies from 2009 to 2023, covering over 5000 firms across more than ten industries, primarily in manufacturing. The sample captures various capacity cycles since the global financial crisis, ensuring strong relevance and timeliness. The selected companies are characterized by relatively dispersed ownership structures, with distinct individuals serving as the chairman and CEO. Additionally, these firms exhibit strong short-term turnover capabilities, growth potential, and innovation capacity, along with robust performance in both management and sales efficiency. Notably, the operational risks associated with these companies are relatively low. Furthermore, our sample spans 34 provinces and 430 cities across China, achieving a diverse regional distribution.

3.2. Variable Description

To better clarify the specific meanings of the key variables and the significance of the measurement indicators, this study presents the category, name, code, and calculation method for each key variable in

Table 1. Variable declaration.

Category	Name	Code	Index
Explained variable	High-quality development of enterprises	TFP_LP	Total factor productivity calculated by Levinsohn–Petrin method
		TFP_OP	Total factor productivity calculated by Olley–Pakes method
		TFP_FE	Total factor productivity calculated by fixed effect method
		TFP_OLS	Total factor productivity calculated by OLS method
Explanatory variable	Diversification of supply chain structure	SCS	Supply chain concentration = (top five suppliers procurement ratio + top five customers sales ratio)/2
		SCS_S	Supplier concentration = the sum of the top five suppliers’ purchases/the total amount of corporate purchases in the year
		SCS_C	Customer concentration = the sum of the sales revenue of the top five customers/the total sales revenue of the enterprise in the year
Mediator variable	Innovation ability	Innovation	Cooperative innovation = ln (number of joint patent applications + 1)
		Knowledge	Knowledge diversification = Teachman entropy index
	Sustainable development ability	ESG_R	According to the ESG rating of ‘C-AAA’ disclosed by the Huazheng index, the values are assigned to 1–9 from high to low.
		ESG_S	The ESG evaluation score disclosed by the Huazheng index
	Risk management ability	Profit	Earnings volatility = (company’s earnings before interest and tax in the past three years/total assets) take the standard deviation
Regulated variable	Supply chain digitization	Digitization	Supply chain digitization = number of supply chain digital keywords/total number of words in the management discussion and analysis section of the annual report × 100
	Operational efficiency	Inventory	Inventory turnover days = ln (365/inventory turnover rate)
		Cash	Cash and cash equivalents turnover ratio = operating income/cash and cash equivalents balance at end of period
	Cultural coordination	Cooperation	Cooperation culture = the number of words of cooperation and its synonyms/the total number of words in the board report section of the annual report
		Confucianism	The influence intensity of Confucian culture = (Max_DIS−DIS)/(Max_DIS−Min_DIS), where Max_DIS and Min_DIS are the maximum and minimum distances from the average geographical distance of all listed companies to the seven Confucian centers in the same year.
	Symmetric information	Information	According to the enterprise information level of ‘ABCD’ disclosed by the Shenzhen Stock Exchange and the Shanghai Stock Exchange, it is assigned to 4–1 from high to low according to ‘excellent-failed’.
		Governance	Internal control index constructed by Dibo database
Control variable	Concentration of power	Duty	If the chairman and the general manager are the same person, the indicator is 1, otherwise it is 0
	Short-term turnover capacity	CAR	Current assets ratio = current assets/total assets
	Innovation potential	IAR	Intangible assets ratio = intangible assets/total assets
	Market value	Tobin	Tobin Q value = market value/total assets
	Development ability	ORG	The growth rate of operating income = (the amount of operating income in the current quarter of the year − the amount of operating income in the previous quarter)/the amount of operating income in the previous quarter
	Management efficiency	OR	Management cost rate = management cost/total assets
	Sale efficiency	SPER	Sales period expense ratio = (sales expense + management expense + financial expense)/(operating income)
	Risk level	OLev	Operating leverage = (net profit + income tax expense + financial expense + depreciation of fixed assets, depreciation of oil and gas assets, depreciation of productive biological assets + amortization of intangible assets + amortization of long-term amortization expenses)/(net profit + income tax expense + financial expense)

, providing a clear reference for readers.

3.2.1. Dependent Variable

The dependent variable in this study is high-quality enterprise development, measured by Total Factor Productivity (TFP). TFP serves as a core quantitative indicator for assessing high-quality development, reflecting the quality, efficiency, and sustainability of development through “input–output efficiency of factors.” Therefore, we adopt TFP as a proxy variable for high-quality development, following the methodologies of Huang et al. (2023) and Dong et al. (2025) [58,103]. Common methods for calculating TFP include Ordinary Least Squares (OLS), Fixed Effects (FE), Generalized Method of Moments (GMM), Olley–Pakes (OP), and Levinsohn–Petrin (LP) methods. The OLS method has a clear calculation logic, requiring only basic panel data on output and factor inputs, making it suitable for preliminary exploratory analysis. The FE method effectively controls for individual fixed heterogeneity, ideal for scenarios with limited factor data. The Olley–Pakes method addresses simultaneity bias and sample selection bias, aligning closely with the actual production behavior of firms. The Levinsohn–Petrin method shares similar estimation logic with Olley–Pakes, but addresses the sample loss issue by accounting for the “non-zero” nature of intermediate inputs. The GMM method can tackle simultaneity bias and control for serial correlation, but its estimation results are highly dependent on the validity of instrumental variables and require a sufficiently long time span for sample data [104]. Consequently, this study employs the LP, OP, FE, and OLS methods to calculate TFP (Supplementary Materials) [105,106,107], represented by the variables TFP_LP, TFP_OP, TFP_FE, and TFP_OLS, respectively. Due to space constraints, the detailed calculation processes of TFP will be reported in the Supplementary Materials for readers’ reference.

3.2.2. Independent Variable

The independent variable in this study is supply chain structure diversification (SCS). The diversification of supply chain structures essentially reflects the manifestation of an enterprise’s “antifragility” at the strategic and operational levels. It requires both forward-looking strategic vision and robust operational capabilities while testing whether the organization can balance efficiency, risk, and social responsibility in complex ecosystems. For enterprises, supply chain diversification reflects their strategic foresight, managerial maturity, and ecosystem mastery, ultimately aiming at building sustainable competitive advantages in uncertain markets. Supply chain concentration is a core inverse indicator of diversification, as it reflects the structural characteristics of supply chains through the dispersion of node distribution. Low concentration indicates higher supply chain diversification, risk dispersion, and responsiveness. Drawing on the measurement approach of Xi et al. (2023) [108], this study uses supply chain concentration to reflect the diversification of an enterprise’s supply chain structure.

Additionally, this study separately analyzes the diversification effects of supply chain structures on high-quality development from upstream and downstream perspectives. Specifically, supplier concentration and customer concentration are used to measure supplier structure (SCS_S) and customer structure (SCS_C), respectively.

3.2.3. Mediating Variable (Medvar)

(1)

Innovation Capability: Referring to Brockman et al. (2018), Xu et al. (2017), and Xu et al. (2025) [109,110,111], this study uses collaborative innovation (Innovation) and knowledge diversification (Knowledge) to represent an enterprise’s innovation capability.

(2)

Sustainability Capability: Following the construction method of Lin et al. (2021) [112], this study uses the ESG rating (ESG_R) and ESG score (ESG_S) from the Huazheng Index to represent sustainability capability.

(3)

Risk Resilience: Drawing on Zhong et al. (2019) [113], this study uses profit volatility (Profit) to measure an enterprise’s risk resilience.

3.2.4. Moderating Variable (Regvar)

In measuring the indicators of supply chain digitalization and collaborative culture, this study employs annual reports from publicly listed companies, text analysis methods, and Python 3.13 technology to assess metrics through keyword frequency analysis. The scientific validity of this approach is supported by four key aspects: First, annual reports are legally mandated documents disclosed by companies in compliance with regulatory requirements, such as those issued by the China Securities Regulatory Commission and stock exchanges, ensuring their authenticity, authority, and comparability, which is critical for providing high-quality data for metric evaluation; Second, the integration of text analysis methods and Python technology effectively transforms the “qualitative descriptions” in annual reports into “quantitative indicators,” facilitating efficient processing and ensuring the objectivity and efficiency of the method from a technical perspective; Third, measuring supply chain digitalization and collaborative culture through keyword frequency is not merely a matter of “counting,” but is grounded in the logic of the “relationship between corporate disclosure tendencies and actual behaviors,” thus providing scientific correlation support; Fourth, a core characteristic of scientific methods is repeatability—other researchers employing the same methodology on the same dataset should yield similar results, and this method fully meets that requirement. Overall, this approach establishes a metric measurement system that aligns with scientific research standards through an “authoritative data foundation, objective technology transformation, associative logic support, and repeatable design assurance.” It facilitates the quantitative characterization of corporate digitalization levels and collaborative culture dimensions while also supporting industry trend analysis and inter-company comparisons, thereby demonstrating substantial scientific and practical value.

(1)

Supply Chain Digitalization (Digitization): This study adopts the approach proposed by Jia et al. (2024) [114], categorizing corporate supply chain digitization into five key dimensions—planning digitalization, procurement digitalization, production digitalization, sales digitalization, and logistics digitalization—based on the *Supply Chain Digitalization Management Guidelines* jointly issued by the State Administration for Market Regulation and the National Standardization Administration in 2022. The primary aim is to construct a relatively objective and comprehensive indicator reflecting the degree of supply chain digitization in Chinese enterprises using text analysis and machine learning methods. The process begins with the construction of a keyword lexicon. To achieve this, the study collected relevant national policy documents and industry research reports published between 2009 and 2023, identifying vocabulary pertinent to corporate supply chain digitization. Additionally, based on the industry classification system established by the China Securities Regulatory Commission in 2012, two representative companies from each industry were selected. Their annual reports were manually reviewed to extract keywords related to supply chain digitization, ensuring a robust initial foundation for the lexicon. To expand the set of similar terms, this study employed the Word2Vec module from the gensim package in Python, training a skip-gram model to represent words as multidimensional vectors and identify similar terms within the keyword set. For each keyword, the five most similar words were selected to further enrich the corporate supply chain digitization terminology dictionary, creating a more comprehensive and objective dataset. We employed a keyword list associated with supply chain digitalization to assess the level of digitalization within the sample enterprises. Detailed information regarding this methodology can be found in the Supplementary Materials. Indicator measurement was conducted by performing text analysis on the “Management Discussion and Analysis” (MD&A) sections of the annual reports of listed companies. The frequency of supply chain digitization-related keywords was counted, and to normalize the data across reports of varying lengths, the total frequency of such keywords was divided by the length of the MD&A section. The resulting value was then multiplied by 100 for convenience, with higher scores indicating a greater degree of supply chain digitization. This approach provides a standardized and quantifiable metric for evaluating corporate supply chain digitization levels across firms.

(2)

Operational Efficiency: Following Zhang et al. (2023) [115], this study uses inventory turnover days (Inventory) and cash and cash equivalents turnover rate (Cash) to measure inventory and cash operational efficiency, respectively.

(3)

Cultural Synergy: This study employs various methods for text analysis, drawing upon the research conducted by Pan et al. (2019) to assess the presence of cooperation within corporate culture through keyword frequency analysis [116]. The specific methodology involves several steps: Initially, this study reviews the vision, mission, and core values of 1000 listed companies to identify terms associated with “cooperation.” Subsequently, these keywords are cross-referenced against the *Chinese Synonyms Dictionary* to identify and supplement any relevant synonyms that may have been overlooked. The final lexicon for text analysis includes terms such as “cooperation, unity, joint, coordination, collaboration, synergy, mutual assistance, sharing, common effort, communication, interaction, and win-win.” To accommodate the linguistic variations across different contexts, synonyms are organized and distinguished by numbering.

The text information utilized for this analysis is drawn from the board reports of annual reports issued by all listed companies between 2009 and 2023. In these board reports, the chairman typically summarizes the company’s past achievements, delineates future development goals, and discusses strategies for attaining those objectives. These narratives often reflect personal values and the methods employed to cultivate a corporate culture that aligns with these goals. Such top-down cultural instillation methods subtly influence employee decision-making, guiding them to address work-related issues in ways that resonate with corporate values.

Consequently, based on the aforementioned cooperation culture lexicon, this study quantifies the occurrences of cooperation and its synonyms within the board reports, dividing this count by the total word count of the relevant section to measure the degree of corporate cooperation culture.

In a complementary approach, this study adopts the methodology of Pan et al. (2022) to evaluate the influence of Confucianism on enterprises through the geographical distance from the company’s registered location to the centers of Confucian culture [117]. Over thousands of years, Confucian culture has established seven cultural centers—Lu, Luo, Shu, Min, Taizhou, Lincun, and East Zhejiang—situated in Qufu (Shandong), Luoyang (Henan), Chengdu (Sichuan), Sanming (Fujian), Dongtai (Jiangsu), Fuzhou (Jiangxi), and Eastern Zhejiang (including Ningbo and Shaoxing). To measure the influence of Confucian culture, the study follows a systematic process: Firstly, it collects the geographical coordinates of the company’s registered location and the Confucian centers using online tools such as Google Maps and Baidu Maps. Secondly, the geographical distances from each registered location to the seven Confucian centers are computed, resulting in the calculation of the average distance (DIS). Finally, the intensity of Confucian cultural influence is quantified using the formula: Confucianism = (Max_DIS − DIS)/(Max_DIS − Min_DIS), where Max_DIS and Min_DIS denote the maximum and minimum average distances from all listed companies to the seven Confucian centers in the same year. A higher value of Confucianism indicates a greater influence of Confucian culture on the enterprise.

(4)

Information Symmetry: Drawing on Xiang and Lu (2020) [118], this study uses the score of information transparency to represent information transparency (Information). Additionally, following Pan et al. (2019) [119], this study employs the internal control index to measure the quality of internal control (Governance).

3.2.5. Control Variables (Control)

To mitigate the potential impact of other factors on the robustness of empirical results, eight control variables (Control) are included: (1) Power Concentration (Duty): Whether the chairman and CEO positions are held by the same individual; (2) Short-term Turnover Ability (CAR): Current asset ratio; (3) Innovation Potential (IAR): Intangible asset ratio; (4) Market Value (Tobin): Tobin’s Q; (5) Growth Ability (ORG): Revenue growth rate; (6) Management Efficiency (OR): Management expense ratio; (7) Sales Efficiency (SPER): Selling and administrative expense ratio; (8) Risk Level (OLev): Operating leverage.

3.3. Model Construction

To examine the impact of supply chain structure diversification on high-quality enterprise development, the following regression models are constructed:

$${\mathrm{T}\mathrm{F}\mathrm{P}}_{\mathrm{i},\mathrm{t}}={\alpha }_0+{\alpha }_1{\mathrm{S}\mathrm{C}\mathrm{S}}_{\mathrm{i},\mathrm{t}}+{\alpha }_i{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i},\mathrm{t}}+\sum \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\sum \mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}+{\mathsf{\epsilon }}_{i,t}$$

(1)

$${\mathrm{T}\mathrm{F}\mathrm{P}}_{\mathrm{i},\mathrm{t}}={\alpha }_0+{\alpha }_1{\mathrm{S}\mathrm{C}\mathrm{S}}_{\mathrm{i},\mathrm{t}}+{\alpha }_i{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i},\mathrm{t}}+\sum \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\sum \mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}+\sum \mathrm{P}\mathrm{r}\mathrm{o}\mathrm{v}\mathrm{i}\mathrm{n}\mathrm{c}\mathrm{e}+\sum \mathrm{C}\mathrm{i}\mathrm{t}\mathrm{y}+{\mathsf{\epsilon }}_{i,t}$$

(2)

In Model (1), i represents the firm, t represents the year, the dependent variable TFP_i,t represents the TFP level of firm i in year t, and the independent variable SCS_i,t represents the degree of supply chain diversification of firm i in year t. Control_i,t includes a series of control variables, and ε is the error term. Industry and Year represent fixed effects for industry and year, respectively. In Model (2), Province represents provincial fixed effects, City additionally represents city fixed effects, while other variables remain unchanged. The coefficient α₁ is the primary focus of this study. Theoretical analysis suggests that if α₁ is significantly positive, it indicates that supply chain structure diversification promotes high-quality development for listed firms, supporting the theoretical expectations of this study.

4. Analysis of Empirical Results

4.1. Descriptive Statistics

Table 2. Summary statistics.

	N	Mean	St.Dev	Min	P25	Median	P75	Max
TFP_LP	43,471	8.385	1.056	6.153	7.645	8.279	9.008	11.238
TFP_OP	43,403	6.728	0.884	4.888	6.105	6.621	7.254	9.156
TFP_FE	43,471	11.216	1.339	8.472	10.266	11.07	12.003	14.896
TFP_OLS	43,471	10.631	1.27	8.012	9.732	10.492	11.377	14.116
SCS	44,829	32.2	17.786	2.89	18.73	29.69	43.2	83.17
SCS_S	39,266	35.654	20.154	5.43	20.36	31.24	46.94	93.85
SCS_C	44,419	33.56	23.262	1.32	15.34	27.74	47.22	97.87
Duty	43,051	0.313	0.464	0	0	0	1	1
CAR	44,854	0.586	0.204	0.105	0.445	0.602	0.745	0.962
IAR	44,854	0.045	0.049	0	0.016	0.032	0.055	0.309
Tobin	42,828	2.04	1.325	0.845	1.252	1.611	2.296	8.804
ORG	43,643	0.355	0.979	−0.725	−0.038	0.12	0.391	6.943
OR	44,849	0.088	0.075	0.008	0.042	0.069	0.109	0.484
SPER	44,849	0.177	0.144	0.015	0.084	0.137	0.219	0.831
OLev	39,745	1.607	0.885	1.015	1.166	1.332	1.665	7.096

presents the descriptive statistics for the 36,344 sample enterprises regarding the primary variables. Both total factor productivity (TFP) and supply chain structure (SCS) exhibit means greater than their medians, indicating a right-skewed distribution. The standard deviations for TFP are 1.056, 0.884, 1.339, and 1.27, which are consistent with previous literature [103]. In contrast, the maximum values for SCS are 83.17, 93.85, and 97.87, while the minimum values are 2.89, 5.43, and 1.32, revealing significant variation among the samples. Overall, it is evident that Chinese listed companies demonstrate strong performance in total factor productivity; However, the excessive concentration in the supply chain connectivity structure of a few sample enterprises may hinder the diversified development of supply chain structures.

4.2. Benchmark Regression

This study utilizes three key indicators, supply chain concentration, supplier concentration, and customer concentration, to assess the level of diversification within corporate supply chain structures. These indicators fundamentally capture the “whole-part” relationship and collectively establish a framework for evaluating supply chain structural diversification. Supplier concentration focuses on the upstream supply side, measuring the extent to which a company depends on a limited number of core suppliers for its procurement activities. Conversely, customer concentration examines the downstream demand side, assessing the degree of reliance on a small group of core customers for sales activities. Supply chain concentration encompasses the entirety of the supply chain, providing a holistic evaluation of a company’s dependence on crucial nodes, effectively integrating and extending the insights gained from the first two indicators. The underlying logic of these concentration measures is that a lower value indicates reduced concentration, reflecting a more dispersed and diversified supply chain structure. As such, concentration acts as a negative core indicator for assessing the degree of supply chain structural diversification. Therefore, in the regression results presented later, if the regression coefficients for the explanatory variables SCS (Supply Chain Concentration), SCS_S (Supplier Concentration), and SCS_C (Customer Concentration) are negative, it signifies that supply chain structural diversification has a positive impact on the high-quality development of the enterprise.

Firstly, based on Model (1), this paper examines the impact of supply chain structure diversification on the high-quality development of enterprises. The empirical results are shown in

Table 3. Baseline regression results.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	TFP_LP		TFP_OP		TFP_FE		TFP_OLS
SCS	−0.015 ***	−0.013 ***	−0.007 ***	−0.005 ***	−0.023 ***	−0.019 ***	−0.021 ***	−0.017 ***
	(−48.299)	(−44.797)	(−25.600)	(−21.495)	(−59.575)	(−52.498)	(−57.091)	(−50.564)
Duty		−0.211 ***		−0.163 ***		−0.299 ***		−0.281 ***
		(−23.544)		(−22.563)		(−26.158)		(−26.017)
CAR		0.171 ***		0.183 ***		−0.804 ***		−0.653 ***
		(6.281)		(8.179)		(−23.161)		(−19.890)
IAR		0.040		−0.234 ***		−0.478 ***		−0.423 ***
		(0.413)		(−2.953)		(−4.002)		(−3.744)
Tobin		−0.072 ***		−0.043 ***		−0.119 ***		−0.110 ***
		(−17.318)		(−12.899)		(−22.025)		(−21.400)
ORG		−0.012 ***		0.007 *		−0.029 ***		−0.025 ***
		(−2.604)		(1.674)		(−4.864)		(−4.402)
OR		−8.170 ***		−6.741 ***		−9.623 ***		−9.279 ***
		(−61.911)		(−61.956)		(−59.068)		(−59.828)
SPER		−0.332 ***		−0.482 ***		−0.588 ***		−0.566 ***
		(−6.115)		(−10.555)		(−8.781)		(−8.891)
OLev		−0.045 ***		−0.053 ***		−0.019 ***		−0.023 ***
		(−9.000)		(−12.736)		(−2.979)		(−3.920)
_cons	7.654 ***	8.779 ***	5.988 ***	6.868 ***	10.549 ***	12.501 ***	9.949 ***	11.757 ***
	(158.468)	(168.037)	(139.842)	(155.670)	(174.533)	(187.722)	(173.788)	(186.936)
N	43,436	36,344	43,368	36,775	43,436	36,344	43,436	36,344
adj. R2	0.173	0.445	0.175	0.467	0.187	0.438	0.185	0.441
Year	yes	yes	yes	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes	yes	yes	yes
r2	0.174	0.446	0.176	0.467	0.188	0.439	0.186	0.442

Note: *** p < 0.01, * p < 0.1.

. As shown in columns (1), (3), (5) and (7), in the regression of fixed annual effect and industry effect, the coefficient of supply chain structure (SCS) is significantly negative at the level of 1%. As shown in columns (2), (4), (6) and (8), the coefficients of supply chain structure (SCS) are still significantly negative at the level of 1% after adding preset control variables. The above results show that on the basis of controlling the annual effect, industry effect and other influencing factors, the diversified development of supply chain structure can significantly promote the high-quality development of enterprises, which verifies the hypothesis H1 of this study.

Similarly, this study examines the impact of upstream supplier structure and downstream customer structure on firms’ high-quality development using Model (1). The empirical results are presented in

Table 4. Based on the supply chain upstream and downstream sub-dimensions.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	TFP_LP		TFP_OP		TFP_FE		TFP_OLS
Column A: Supplier structure and high-quality development of enterprises
SCS_S	−0.010 ***	−0.008 ***	−0.004 ***	−0.002 ***	−0.016 ***	−0.013 ***	−0.015 ***	−0.012 ***
	(−35.760)	(−32.937)	(−14.573)	(−9.548)	(−46.780)	(−41.739)	(−44.252)	(−39.502)
Control		yes		yes		yes		yes
_cons	7.683 ***	8.905 ***	5.951 ***	6.845 ***	10.582 ***	12.706 ***	9.972 ***	11.934 ***
	(90.919)	(110.352)	(77.559)	(101.706)	(103.173)	(125.474)	(102.009)	(124.676)
N	37,928	31,528	37,933	31,935	37,928	31,528	37,928	31,528
adj. R2	0.163	0.431	0.176	0.460	0.172	0.423	0.171	0.426
Year	yes	yes	yes	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes	yes	yes	yes
r2	0.163	0.432	0.177	0.461	0.173	0.424	0.172	0.427
Column B: Customer structure and high-quality development of enterprises
SCS_C	−0.010 ***	−0.008 ***	−0.005 ***	−0.004 ***	−0.015 ***	−0.012 ***	−0.014 ***	−0.011 ***
	(−43.889)	(−38.368)	(−28.354)	(−23.160)	(−50.438)	(−42.347)	(−49.020)	(−41.344)
Control		yes		yes		yes		yes
_cons	7.692 ***	8.832 ***	6.028 ***	6.910 ***	10.586 ***	12.562 ***	9.986 ***	11.816 ***
	(155.733)	(168.102)	(138.523)	(155.797)	(170.802)	(187.577)	(170.169)	(186.842)
N	43,032	35,999	42,967	36,430	43,032	35,999	43,032	35,999
adj. R2	0.164	0.435	0.180	0.469	0.166	0.420	0.166	0.425
Year	yes	yes	yes	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes	yes	yes	yes
r2	0.165	0.436	0.181	0.470	0.166	0.421	0.167	0.426

Note: *** p < 0.01.

. As shown in columns (1), (3), (5), and (7), the coefficients of supplier structure (SCS_S) and customer structure (SCS_C) are significantly negative at the 1% level in regressions controlling for year and industry fixed effects. Likewise, as shown in columns (2), (4), (6), and (8), after incorporating the pre-specified control variables, the coefficients of supplier structure (SCS_S) and customer structure (SCS_C) remain significantly negative at the 1% level. These findings further demonstrate that, after controlling for year effects, industry effects, and other influencing factors, the diversification of supply chain structure significantly promotes firms’ high-quality development, which also confirms hypothesis H1 of this study.

4.3. Mechanism Verification

Based on the theoretical analysis and benchmark regression results above, this study further explores the mechanisms through which supply chain structure diversification influences firms’ high-quality development. Diversification in supply chain structure can enhance firms’ innovation capacity, sustainability, and risk resilience, thereby promoting high-quality development.

Following the mediation testing method proposed by Wen and Ye (2014) [120], this study defines the mechanism test variable (Medvar) and constructs regression models (4) and (5) to examine the coefficients of key variables across Models (1), (4), and (5).

$${\mathrm{M}\mathrm{e}\mathrm{d}\mathrm{v}\mathrm{a}\mathrm{r}}_{\mathrm{i},\mathrm{t}}={\beta }_0+{\beta }_1{\mathrm{C}\mathrm{C}\mathrm{R}}_{\mathrm{i},\mathrm{t}}+{\beta }_i{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i},\mathrm{t}}+\sum \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\sum \mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}+{\mathsf{\epsilon }}_{i,t}$$

(3)

$${\mathrm{T}\mathrm{F}\mathrm{P}}_{\mathrm{i},\mathrm{t}}={\mathrm{\delta }}_0+{\mathrm{\delta }}_1{\mathrm{S}\mathrm{C}\mathrm{S}}_{\mathrm{i},\mathrm{t}}+{\mathrm{\delta }}_2{\mathrm{M}\mathrm{e}\mathrm{d}\mathrm{v}\mathrm{a}\mathrm{r}}_{i,t}+{\mathrm{\delta }}_i{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i},\mathrm{t}}+\sum \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\sum \mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}+{\mathsf{\epsilon }}_{i,t}$$

(4)

Table 5. The mediating effect of innovation ability.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
	Innovation	TFP_LP	TFP_OP	TFP_FE	TFP_OLS	Knowledge	TFP_LP	TFP_OP	TFP_FE	TFP_OLS
Column A: Supply chain structure, innovation ability and high-quality development of enterprises
SCS	−0.009 ***	−0.011 ***	−0.004 ***	−0.016 ***	−0.015 ***	−0.007 ***	−0.011 ***	−0.004 ***	−0.016 ***	−0.014 ***
	(−28.722)	(−38.348)	(−15.626)	(−45.774)	(−43.850)	(−24.698)	(−34.670)	(−16.119)	(−40.121)	(−38.615)
Innovation		0.196 ***	0.140 ***	0.270 ***	0.254 ***
		(48.340)	(44.009)	(51.629)	(51.292)
Knowledge							0.386 ***	0.256 ***	0.532 ***	0.498 ***
							(60.366)	(49.539)	(64.244)	(63.631)
Year	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Column B: Supplier structure, innovation ability and high-quality development of enterprises
SCS_S	−0.008 ***	−0.007 ***	−0.001 ***	−0.011 ***	−0.010 ***	−0.007 ***	−0.006 ***	−0.001 ***	−0.010 ***	−0.009 ***
	(−29.856)	(−26.378)	(−3.703)	(−34.799)	(−32.586)	(−28.587)	(−22.152)	(−2.763)	(−28.325)	(−26.521)
Innovation		0.196 ***	0.143 ***	0.270 ***	0.254 ***
		(45.154)	(42.493)	(47.949)	(47.710)
Knowledge							0.383 ***	0.260 ***	0.525 ***	0.493 ***
							(54.714)	(47.098)	(57.977)	(57.538)
Year	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Column C: Customer structure, innovation ability and high-quality development of enterprises
SCS_C	−0.005 ***	−0.007 ***	−0.003 ***	−0.010 ***	−0.009 ***	−0.003 ***	−0.007 ***	−0.004 ***	−0.010 ***	−0.009 ***
	(−18.119)	(−34.478)	(−19.418)	(−38.448)	(−37.434)	(−13.385)	(−32.468)	(−21.017)	(−35.229)	(−34.528)
Innovation		0.205 ***	0.141 ***	0.286 ***	0.268 ***
		(51.047)	(44.825)	(54.902)	(54.419)
Knowledge							0.403 ***	0.260 ***	0.558 ***	0.521 ***
							(63.400)	(50.910)	(67.576)	(66.875)
Year	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes

Note: *** p < 0.01.

,

Table 6. The mediating effect of sustainable development ability.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
	ESG_R	TFP_LP	TFP_OP	TFP_FE	TFP_OLS	ESG_S	TFP_LP	TFP_OP	TFP_FE	TFP_OLS
Column A: Supply chain structure, sustainable development ability and high-quality development of enterprises
SCS	−0.007 ***	−0.012 ***	−0.005 ***	−0.018 ***	−0.016 ***	−0.036 ***	−0.012 ***	−0.005 ***	−0.018 ***	−0.016 ***
	(−21.305)	(−42.638)	(−20.231)	(−49.691)	(−47.850)	(−23.230)	(−42.322)	(−19.986)	(−49.320)	(−47.487)
ESG_R		0.076 ***	0.042 ***	0.125 ***	0.115 ***
		(16.730)	(11.433)	(21.487)	(20.815)
ESG_S							0.017 ***	0.010 ***	0.028 ***	0.026 ***
							(18.225)	(12.668)	(23.103)	(22.409)
Year	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Column B: Supplier structure, sustainable development ability and high-quality development of enterprises
SCS_S	−0.005 ***	−0.008 ***	−0.002 ***	−0.013 ***	−0.012 ***	−0.024 ***	−0.008 ***	−0.002 ***	−0.013 ***	−0.011 ***
	(−16.378)	(−31.659)	(−9.012)	(−39.886)	(−37.735)	(−17.682)	(−31.403)	(−8.805)	(−39.600)	(−37.453)
ESG_R		0.078 ***	0.043 ***	0.130 ***	0.119 ***
		(16.140)	(11.184)	(20.790)	(20.131)
ESG_S							0.018 ***	0.010 ***	0.030 ***	0.027 ***
							(17.881)	(12.667)	(22.659)	(21.978)
Year	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Column C: Customer structure, sustainable development ability and high-quality development of enterprises
SCS_C	−0.005 ***	−0.008 ***	−0.004 ***	−0.011 ***	−0.010 ***	−0.024 ***	−0.008 ***	−0.004 ***	−0.011 ***	−0.010 ***
	(−18.169)	(−36.669)	(−22.433)	(−39.980)	(−39.085)	(−20.179)	(−36.348)	(−22.201)	(−39.591)	(−38.706)
ESG_R		0.082 ***	0.042 ***	0.136 ***	0.124 ***
		(17.806)	(11.524)	(22.833)	(22.094)
ESG_S							0.019 ***	0.010 ***	0.031 ***	0.028 ***
							(19.334)	(12.710)	(24.509)	(23.741)
Year	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes	yes	yes	yes	yes	yes

Note: *** p < 0.01.

and

Table 7. The mediating effect of risk resilience.

	(1)	(2)	(3)	(4)	(5)
	Profit	TFP_LP	TFP_OP	TFP_FE	TFP_OLS
Column A: Supply chain structure, risk resilience and high-quality development of enterprises
SCS	0.031 ***	−0.012 ***	−0.005 ***	−0.018 ***	−0.017 ***
	(19.213)	(−43.893)	(−21.094)	(−51.041)	(−49.182)
Profit		−0.008 ***	−0.002 ***	−0.017 ***	−0.015 ***
		(−7.174)	(−2.717)	(−12.315)	(−11.547)
Year	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes
Column B: Supplier structure, risk resilience and high-quality development of enterprises
SCS_S	0.023 ***	−0.008 ***	−0.002 ***	−0.013 ***	−0.012 ***
	(16.299)	(−31.989)	(−9.042)	(−40.366)	(−38.184)
Profit		−0.009 ***	−0.004 ***	−0.018 ***	−0.017 ***
		(−7.854)	(−3.750)	(−12.730)	(−12.004)
Year	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes
Column C: Customer structure, risk resilience and high-quality development of enterprises
SCS_C	0.019 ***	−0.008 ***	−0.004 ***	−0.011 ***	−0.010 ***
	(16.050)	(−37.556)	(−22.874)	(−41.059)	(−40.126)
Profit		−0.009 ***	−0.003 ***	−0.019 ***	−0.017 ***
		(−8.335)	(−2.865)	(−13.743)	(−12.908)
Year	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes

Note: *** p < 0.01.

present the results of mechanism tests for the impact of supply chain structure diversification on firms’ high-quality development, analyzed as follows.

4.3.1. Mediating Effect of Innovation Capability

Table 5 reports the mechanism test results for the impact of supply chain structure diversification on firms’ high-quality development through “innovation capability.” In columns (1) and (6), the regression coefficients of supply chain structure (SCS), supplier structure (SCS_S), and customer structure (SCS_C) are all significantly negative at the 1% level, confirming a positive relationship between supply chain structure diversification and firms’ “innovation capability.” In columns (2)–(5) and (7)–(10), the regression coefficients of SCS, SCS_S, and SCS_C remain significantly negative at the 1% level, while the coefficients of cooperative innovation (Innovation) and knowledge diversification (Knowledge) are significantly positive at the 1% level. For robustness, a Sobel test was conducted to examine the presence of a significant mediating effect, and the results confirm the existence of the mediation effect through the “innovation capability” channel. These findings indicate that supply chain structure diversification enhances firms’ innovation capabilities, thereby promoting high-quality development. This provides empirical support for Hypothesis H2 of this study.

4.3.2. Mediating Effect of Sustainability Capability

Table 6 presents the mechanism test results for the impact of supply chain structure diversification on firms’ high-quality development through “sustainability capability.” In columns (1) and (6), the regression coefficients of supply chain structure (SCS), supplier structure (SCS_S), and customer structure (SCS_C) are all significantly negative at the 1% level, confirming a positive relationship between supply chain structure diversification and firms’ “sustainability capability.” In columns (2)–(5) and (7)–(10), the regression coefficients of SCS, SCS_S, and SCS_C remain significantly negative at the 1% level, while the coefficients of ESG ratings (ESG_R) and ESG scores (ESG_S) are significantly positive at the 1% level. To ensure robustness, a Sobel test was conducted, and the results confirm the significant mediating effect of the “sustainability capability” channel. These findings suggest that supply chain structure diversification enhances firms’ sustainability capabilities, thereby promoting high-quality development. This provides empirical support for Hypothesis H3.

4.3.3. Mediating Effect of Risk Management Capability

Table 7 provides the mechanism test results for the impact of supply chain structure diversification on firms’ high-quality development through “risk management capability.” In column (1), the regression coefficients of supply chain structure (SCS), supplier structure (SCS_S), and customer structure (SCS_C) are all significantly positive at the 1% level, confirming a positive relationship between supply chain structure diversification and firms’ “risk management capability.” In columns (2)–(5), the regression coefficients of SCS, SCS_S, and SCS_C remain significantly negative at the 1% level, while the coefficients of profit volatility (Profit) are significantly negative at the 1% level. For robustness, a Sobel test was conducted, confirming the significant mediating effect of the “risk management capability” channel. These results indicate that supply chain structure diversification strengthens firms’ risk management capabilities, thereby promoting high-quality development. This provides empirical support for Hypothesis H4.

4.4. Robustness Test

4.4.1. Replacing the Independent Variables

To further validate the robustness of the findings, this study follows the methodology of Lin and Wei (2023) by replacing the independent variables supply chain structure (SCS) [121], supplier structure (SCS_S), and customer structure (SCS_C) with dummy variables: M.SCS, M.SCS_S, and M.SCS_C. Specifically, if SCS is above the industry’s annual median, the supply chain structure is considered highly diversified, and the dummy variable M.SCS is set to 1; otherwise, it is set to 0. Similarly, dummy variables were created for supplier structure (M.SCS_S) and customer structure (M.SCS_C). The models were re-estimated accordingly, and the empirical results are shown in Column A of

Table 8. Considering the influence of core explanatory variables and other supply chain behaviors.

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

TFP_LP

TFP_OP

TFP_FE

TFP_OLS

Column A: Replace explanatory variables

M.SCS

−0.373 ***

−0.171 ***

−0.541 ***

−0.497 ***

(−42.216)

(−24.011)

(−47.797)

(−46.424)

M.SCS_S

−0.273 ***

−0.083 ***

−0.415 ***

−0.375 ***

(−30.943)

(−11.699)

(−36.410)

(−34.891)

M.SCS_C

−0.336 ***

−0.184 ***

−0.465 ***

−0.431 ***

(−38.181)

(−25.762)

(−41.240)

(−40.512)

Year

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Industry

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Column B: Increase supply chain related control variables

SCS

−0.012 ***

−0.005 ***

−0.018 ***

−0.016 ***

(−41.100)

(−20.186)

(−47.168)

(−45.535)

SCS_S

−0.008 ***

−0.002 ***

−0.012 ***

−0.011 ***

(−29.219)

(−7.972)

(−36.720)

(−34.743)

SCS_C

−0.008 ***

−0.004 ***

−0.011 ***

−0.010 ***

(−36.100)

(−22.470)

(−38.860)

(−38.062)

Control

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Risk

0.254 ***

0.271 ***

0.255 ***

0.222 ***

0.229 ***

0.220 ***

0.312 ***

0.337 ***

0.316 ***

0.300 ***

0.324 ***

0.304 ***

(14.988)

(14.830)

(14.962)

(16.203)

(15.655)

(16.106)

(14.314)

(14.373)

(14.281)

(14.622)

(14.622)

(14.588)

Quality

−0.186 **

−0.093

−0.206 ***

−0.227 ***

−0.225 ***

−0.227 ***

−0.205 **

−0.072

−0.241 **

−0.203 **

−0.084

−0.235 ***

(−2.425)

(−1.034)

(−2.721)

(−3.581)

(−3.124)

(−3.615)

(−2.144)

(−0.648)

(−2.541)

(−2.246)

(−0.798)

(−2.619)

Finance

0.208 ***

0.213 ***

0.222 ***

0.147 ***

0.155 ***

0.152 ***

0.261 ***

0.265 ***

0.282 ***

0.248 ***

0.253 ***

0.267 ***

(29.306)

(28.470)

(31.213)

(25.629)

(26.100)

(26.653)

(28.369)

(27.365)

(30.335)

(28.526)

(27.595)

(30.467)

IME

0.046 ***

0.068 ***

0.045 ***

0.032 ***

0.044 ***

0.033 ***

0.026 **

0.046 ***

0.023 *

0.029 **

0.049 ***

0.027 **

(4.381)

(5.924)

(4.376)

(3.418)

(4.507)

(3.602)

(2.104)

(3.344)

(1.946)

(2.490)

(3.737)

(2.368)

Year

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Industry

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Note: *** p < 0.01, ** p < 0.05, * p < 0.1.

.

The regression coefficients of the dummy variables M.SCS, M.SCS_S, and M.SCS_C are all significantly negative at the 1% level, consistent with the previous results, confirming the robustness of the conclusions.

4.4.2. Controlling for Other Supply Chain Factors

Supply chain risk serves as the primary hedging variable for supply chain structural diversification, playing a crucial role in determining the stability of a company’s high-quality development. Customer quality acts as a value filter for supply chain diversity, significantly impacting the profitability associated with high-quality development. Supply chain finance functions as a liquidity lubricant for diversification, influencing the liquidity available for facilitating the enterprise’s high-quality growth. Additionally, supply chain efficiency serves as an operational benchmark, determining the cost competitiveness of the high-quality development process. These four factors do not function independently; Rather, they interact synergistically to shape the overall impact of supply chain structural diversification on promoting high-quality development within enterprises. Together, they create a comprehensive framework that underscores the importance of each element in achieving sustainable growth and competitive advantage.

To exclude the influence of other supply chain-related factors on corporate decision-making and development quality, additional variables were incorporated into the model while maintaining the original control variables. This adjustment aims to explore the net effect of supply chain structure diversification on high-quality development. Supply Chain Risk: Measured by the deviation between production and demand volatility (Risk). Supply Chain Quality of Customers: Measured by the proportion of listed companies among the top five customers (Quality). Supply Chain Finance: Measured by the frequency of supply chain finance-related keywords in annual reports using textual analysis (Finance). Supply Chain Efficiency: Measured by inventory turnover (IME).

The empirical regression results, shown in Column B of Table 8, indicate that the coefficients of SCS, SCS_S, and SCS_C remain significantly negative at the 1% level, consistent with the earlier findings. These results confirm that the conclusions are robust.

4.4.3. Adding Regional Fixed Effects

Industry and year fixed effects can only control for common industry characteristics and macroeconomic shocks; However, they do not account for unique omitted variables at the regional level. To address this gap and enhance the precision of estimates as well as the reliability of conclusions, this study incorporates province and city fixed effects in the following sections.

To further verify the robustness of the results, province fixed effects and city fixed effects were added to regression model (2) while continuing to control for industry, year, and province fixed effects.

The empirical results, shown in

Table 9. Increase the fixed effect.

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

TFP_LP

TFP_OP

TFP_FE

TFP_OLS

SCS

−0.012 ***

−0.005 ***

−0.018 ***

−0.017 ***

(−43.512)

(−21.164)

(−51.131)

(−49.265)

SCS_S

−0.008 ***

−0.002 ***

−0.013 ***

−0.012 ***

(−32.531)

(−10.151)

(−41.095)

(−38.957)

SCS_C

−0.008 ***

−0.004 ***

−0.011 ***

−0.010 ***

(−36.993)

(−22.616)

(−40.811)

(−39.877)

Year

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Industry

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Province

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

City

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Note: *** p < 0.01.

, reveal that the regression coefficients of SCS, SCS_S, and SCS_C remain significantly negative at the 1% level. These findings are consistent with earlier results, further confirming the robustness of the conclusions.

4.4.4. Excluding the Impact of the COVID-19 Pandemic

Since 2019, the COVID-19 pandemic, widely regarded as a major global public health crisis, has had profound and far-reaching effects on economic development and public safety. The urgent adjustments required by the pandemic, coupled with external constraints, have subjected enterprises to triple pressures: “rising costs, coordination difficulties, and resource dispersion,” potentially pushing them into a diversification trap. Additionally, the pandemic has introduced significant cash flow pressures, supply chain disruptions, and heightened external uncertainties, all of which have directly constrained the resources available for high-quality development and, in some cases, resulted in deviations from intended transformation trajectories. Given the significant impact of the pandemic, this study excludes research samples from the period 2020 to 2023. Instead, the sample time window is adjusted to cover 2009 to 2019. An empirical regression analysis is then conducted to reassess the findings within this adjusted timeframe.

The regression results, presented in Column A of

Table 10. Exclude the impact of events and industries.

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

TFP_LP

TFP_OP

TFP_FE

TFP_OLS

Column A: Exclude the impact of the new coronavirus pneumonia epidemic

SCS

−0.012 ***

−0.005 ***

−0.018 ***

−0.016 ***

(−33.124)

(−16.762)

(−38.403)

(−37.102)

SCS_S

−0.008 ***

−0.002 ***

−0.013 ***

−0.011 ***

(−24.394)

(−8.158)

(−30.546)

(−29.018)

SCS_C

−0.008 ***

−0.004 ***

−0.011 ***

−0.010 ***

(−28.467)

(−16.588)

(−31.387)

(−30.618)

Year

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Industry

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Column B: Exclude industry influence

SCS

−0.013 ***

−0.006 ***

−0.019 ***

−0.017 ***

(−38.889)

(−22.760)

(−43.145)

(−41.925)

SCS_S

−0.010 ***

−0.003 ***

−0.014 ***

−0.013 ***

(−30.406)

(−10.320)

(−36.351)

(−34.600)

SCS_C

−0.008 ***

−0.005 ***

−0.011 ***

−0.010 ***

(−30.475)

(−23.915)

(−31.772)

(−31.438)

Year

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Industry

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Note: *** p < 0.01.

, show that the coefficients of supply chain structure (SCS), supplier structure (SCS_S), and customer structure (SCS_C) remain significantly negative at the 1% level. These findings are consistent with prior results, indicating the robustness of the conclusions.

4.4.5. Excluding Industry Effects

Given that the purchase-production-sales processes and supply chain systems vary significantly across industries, the role of supply chain structure in promoting high-quality development may differ as well. Manufacturing, for example, is a capital- and technology-intensive sector requiring substantial investment in procurement, R&D, production, and sales, with a strong need for coordination and collaboration across these processes. Based on this, and in addition to controlling for industry factors, this study focuses solely on listed manufacturing firms by excluding other industries from the sample and re-estimates the model.

The regression results, shown in Column B of Table 10, indicate that the coefficients of SCS, SCS_S, and SCS_C remain significantly negative at the 1% level. These results are consistent with previous findings, further confirming the robustness of the conclusions.

4.5. Endogeneity Test

4.5.1. Lagged Variable Approach

In this study, the primary value of employing lagged treatment is its capacity to restore the economic reality over time. This approach effectively tackles the fundamental challenge of endogeneity in causal identification, while also accurately capturing the dynamic effects of supply chain structural diversification on high-quality enterprise development. Ultimately, this enhances both the causal interpretability and policy relevance of the regression results. To mitigate potential concerns regarding reverse causality and temporal lag effects, this paper applies lagged treatment to both the dependent and independent variables, following the framework established in Model (1).

On one hand, the study lags the dependent variable, high-quality corporate development (TFP), by one period. The regression results, presented in Column A of

Table 11. Lag processing of main variables.

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

TFP_LP

TFP_OP

TFP_FE

TFP_OLS

Column A: Lag processing of high-quality development of enterprises

SCS

−0.012 ***

−0.005 ***

−0.019 ***

−0.017 ***

(−39.178)

(−17.902)

(−46.846)

(−44.982)

SCS_S

−0.008 ***

−0.002 ***

−0.013 ***

−0.012 ***

(−29.275)

(−8.254)

(−37.645)

(−35.576)

SCS_C

−0.008 ***

−0.004 ***

−0.011 ***

−0.011 ***

(−33.650)

(−19.411)

(−37.847)

(−36.836)

Year

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Industry

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Column B: Lag processing of supply chain structure diversification

L.SCS

−0.013 ***

−0.005 ***

−0.019 ***

−0.017 ***

(−42.240)

(−21.050)

(−49.062)

(−47.360)

L.SCS_S

−0.008 ***

−0.002 ***

−0.013 ***

−0.012 ***

(−30.860)

(−9.558)

(−38.650)

(−36.652)

L.SCS_C

−0.008 ***

−0.004 ***

−0.012 ***

−0.011 ***

(−36.163)

(−22.305)

(−39.644)

(−38.761)

Year

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Industry

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Note: *** p < 0.01.

, show that the coefficients of supply chain structure (SCS), supplier structure (SCS_S), and customer structure (SCS_C) remain significantly negative at the 1% level. These findings are highly consistent with previous results, further supporting the robustness of the conclusions.

On the other hand, this study lags the independent variable, supply chain structure (SCS), by one period. The regression results, presented in Column B of Table 11, indicate that the coefficients of supply chain structure (SCS), supplier structure (SCS_S), and customer structure (SCS_C) remain significantly negative at the 1% level. These findings are highly consistent with previous results. This evidence alleviates potential endogeneity concerns to some extent.

4.5.2. Heckman Two-Step Test

In the process of advancing high-quality development through supply chain structural diversification, the core function of the Heckman two-step method is to address sample selection bias resulting from “self-selection of enterprises into supply chain structural diversification.” This approach involves two modeling steps: First capturing the selection process of supply chain diversification, and then controlling for unobserved self-selection characteristics using the inverse Mills ratio. Ultimately, this method accurately estimates the true causal effect of supply chain structural diversification on high-quality enterprise development.

To address potential endogeneity caused by selection bias in core variable choices, this study employs the Heckman two-step method for re-estimation. In the first stage, firm guild culture, industry Lerner index, and environmental subsidies are used as control variables, applying a Probit model with the same control variables as in Model (1). The inverse Mills ratio (IMR) obtained from the first step is then incorporated into the second-stage model to control for endogeneity due to selection bias. The empirical regression results, shown in Column A of

Table 12. Heckman two-stage test and instrumental variable method.

Variables	First Stage		Second Stage
	(1)	(2)	(3)	(4)	(5)	(6)
	SCS	SCS_D	TFP_LP	TFP_OP	TFP_FE	TFP_OLS
Column A: Heckman two-stage test
SCS			−0.010 ***	−0.003 ***	−0.014 ***	−0.013 ***
			(0.000)	(0.000)	(0.000)	(0.000)
IMR			−15.350 ***	−10.677 ***	−18.606 ***	−17.667 ***
			(1.369)	(1.121)	(1.733)	(1.641)
_cons		−11.693 ***	9.828 ***	7.852 ***	13.760 ***	12.979 ***
		(3.982)	(0.029)	(0.023)	(0.036)	(0.034)
N		28,980	25,785	26,024	25,785	25,785
R2			0.360	0.381	0.378	0.379
Column B: Instrumental variable method
IV.SCS	0.008 ***
	(15.556)
SCS			−0.001 *	−0.002 ***	−0.002 ***	−0.002 ***
			(−0.521)	(−4.551)	(−2.673)	(−2.653)
FE	yes		yes	yes	yes	yes
Wald F	7240.1
Sargan	0.000

Note: *** p < 0.01, * p < 0.1.

, indicate that after applying the Heckman two-step method, supply chain structure diversification still has a significant positive impact on high-quality corporate development. This alleviates potential endogeneity concerns in the study.

4.5.3. Instrumental Variable Method

In this study, the core value of the instrumental variable method lies in overcoming the constraints of endogeneity to achieve precise identification of causal relationships. By introducing exogenous instrumental variables, it isolates the endogenous variation in supply chain structural diversification caused by reverse causality, omitted variables, and measurement errors, thereby revealing the net effect of “supply chain structural diversification on promoting high-quality development.”

To further address potential endogeneity issues, this study adopts the instrumental variable (IV) method. The industry-level average of supply chain structure (IV.SCS) is selected as an instrumental variable for supply chain structure diversification. Based on this, the study utilizes the two-stage least squares (2SLS) method for testing. The regression results, presented in Column B of Table 12, show that in the first-stage regression, the industry average supply chain structure significantly promotes a firm’s supply chain structure diversification. The Wald F-statistic is 7240.1, which exceeds the 10% weak identification critical value of 16.38, indicating no weak instrument problem. In the second-stage regression, supply chain structure diversification continues to have a positive impact on high-quality corporate development. These findings are consistent with previous results and further mitigate potential endogeneity concerns.

4.6. Moderating Effect Test

Based on the theoretical analysis and baseline regression results, this study further explores the moderating effect of supply chain digitization, operational efficiency, cultural synergy, and information symmetry in the relationship between supply chain structure diversification and high-quality corporate development. Supply chain digitization, operational efficiency, cultural synergy, and information symmetry serve as a significant positive driver in enhancing the impact of supply chain structure diversification on high-quality corporate development. To examine this, a moderating variable (Regvar) and its interaction term with supply chain structure (SCS) are incorporated into Model (1), constructing Model (5) to test the regression coefficients of key variables:

$${\mathrm{T}\mathrm{F}\mathrm{P}}_{\mathrm{i},\mathrm{t}}={\mathrm{\eta }}_0+{\mathrm{\eta }}_1{\mathrm{S}\mathrm{C}\mathrm{S}}_{i,t}+{\mathrm{\eta }}_2{\mathrm{S}\mathrm{C}\mathrm{S}}_{i,t}\times {\mathrm{R}\mathrm{e}\mathrm{g}\mathrm{v}\mathrm{a}\mathrm{r}}_{i,t}+{\mathrm{\eta }}_3{\mathrm{R}\mathrm{e}\mathrm{g}\mathrm{v}\mathrm{a}\mathrm{r}}_{i,t}+{\mathrm{\eta }}_i{\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}}_{\mathrm{i},\mathrm{t}}+\sum \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}+\sum \mathrm{I}\mathrm{n}\mathrm{d}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{r}\mathrm{y}+{\mathsf{\epsilon }}_{i,t}$$

(5)

4.6.1. The Impact of Supply Chain Digitization

In the process of supply chain structure diversification promoting high-quality corporate development, supply chain digitization acts as a core engine driving industrial upgrading by reconstructing resource allocation logic, fostering innovation, reshaping sustainable development models, and enhancing risk resistance. When supply chain digitization deeply integrates with diversification strategies, firms can establish a “digital immune system,” enabling sustainable high-quality development within complex business environments.

The moderating effect test results are reported in Columns (1), (4), (7), and (10) of

Table 13. The impact of supply chain digitization on operational efficiency.

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

(12)

TFP_LP

TFP_OP

TFP_FE

TFP_OLS

SCS

−0.013 ***

−0.013 ***

−0.013 ***

−0.005 ***

−0.006 ***

−0.005 ***

−0.019 ***

−0.019 ***

−0.019 ***

−0.017 ***

−0.018 ***

−0.018 ***

(−42.727)

(−47.811)

(−47.163)

(−20.407)

(−24.904)

(−23.245)

(−50.940)

(−54.582)

(−54.736)

(−48.957)

(−52.794)

(−52.827)

Digitization

0.386 ***

0.199 ***

0.324 ***

0.316 ***

(8.174)

(5.133)

(5.562)

(5.707)

Inventory

−0.121 ***

−0.083 ***

−0.118 ***

−0.116 ***

(−25.011)

(−19.620)

(−19.758)

(−20.339)

Cash

0.024 ***

0.020 ***

0.027 ***

0.026 ***

(37.949)

(37.771)

(34.486)

(35.186)

SCS × Digitization

−0.013 ***

−0.008 **

−0.027 ***

−0.024 ***

(−3.239)

(−2.523)

(−5.625)

(−5.292)

SCS × Inventory

−0.001 ***

−0.001 ***

−0.001 ***

−0.001 ***

(−2.829)

(−4.456)

(−2.671)

(−2.864)

SCS × Cash

−0.016 ***

−0.015 ***

−0.026 ***

−0.024 ***

(−4.159)

(−4.702)

(−5.285)

(−5.194)

Year

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Industry

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

yes

Note: *** p < 0.01, ** p < 0.05.

. The regression coefficients of supply chain structure (SCS) are significantly negative at the 1% level. The interaction term coefficients (SCS × Digitization) are significantly negative at the 1%, 5%, 1%, and 1% levels, respectively. These results suggest that firms with higher levels of supply chain digitization experience a more pronounced impact of supply chain structure diversification on high-quality corporate development.

4.6.2. The Impact of Operational Efficiency

Inventory turnover days as well as cash and cash equivalents turnover ratio are critical financial indicators for measuring a firm’s operational efficiency. Together, they provide a comprehensive evaluation of the firm’s efficiency across the entire value chain, from procurement, production, and sales to cash collection, helping managers identify operational bottlenecks and optimize resource allocation. The heterogeneity test results of Inventory Turnover Days are reported in Columns (2), (5), (8), and (11) of Table 13. The regression coefficients of supply chain structure (SCS) and the interaction term (SCS × Inventory) are significantly negative at the 1% level. The heterogeneity test results of Cash and Cash Equivalents Turnover Ratio are reported in Columns (3), (6), (9), and (12) of Table 13. The regression coefficients of supply chain structure (SCS) and the interaction term (SCS × Cash) are also significantly negative at the 1% level.

These findings suggest that firms with higher operational efficiency experience a more pronounced positive impact of supply chain structure diversification on high-quality corporate development.

4.6.3. The Impact of Cultural Synergy

The resonance between cooperative culture and Confucian culture represents a deep integration of “relational rationality” from Chinese wisdom and “instrumental rationality” of modern supply chains. The former provides emotional connections and ethical constraints, while the latter builds an efficiency framework and technical support. This “unity of reason and emotion” cultural system enables diverse supply chains to maintain ecological diversity while establishing a sense of internal order. This ultimately fosters a high-quality development paradigm capable of resisting global risks and sustaining innovative vitality. Column A of

Table 14. The influence of cultural synergy and information symmetry.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	TFP_LP		TFP_OP		TFP_FE		TFP_OLS
Column A: Heterogeneity Analysis of Cultural Symmetry
SCS	−0.013 ***	−0.013 ***	−0.005 ***	−0.005 ***	−0.019 ***	−0.019 ***	−0.017 ***	−0.017 ***
	(−44.902)	(−44.497)	(−21.433)	(−21.326)	(−52.720)	(−52.206)	(−50.760)	(−50.282)
Cooperation	139.743 ***		110.060 ***		165.243 ***		158.480 ***
	(12.065)		(11.573)		(11.338)		(11.478)
Confucianism		−0.248 ***		−0.155 ***		−0.355 ***		−0.331 ***
		(−7.097)		(−5.258)		(−7.878)		(−7.772)
SCS × Cooperation	−5.617 ***		−3.400 ***		−8.409 ***		−7.790 ***
	(−7.547)		(−5.545)		(−8.906)		(−8.718)
SCS × Confucianism		−0.007 ***		−0.008 ***		−0.007 **		−0.007 **
		(−2.789)		(−3.462)		(−2.236)		(−2.339)
Year	yes	yes	yes	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes	yes	yes	yes
Column B: Heterogeneity analysis of information symmetry
SCS	−0.011 ***	−0.010 ***	−0.004 ***	−0.003 ***	−0.016 ***	−0.016 ***	−0.015 ***	−0.014 ***
	(−35.353)	(−33.442)	(−14.642)	(−12.669)	(−42.766)	(−40.558)	(−40.901)	(−38.750)
Information	0.232 ***		0.135 ***		0.349 ***		0.322 ***
	(27.865)		(19.955)		(32.804)		(32.006)
Governance		9.759 ***		6.464 ***		13.699 ***		12.815 ***
		(21.532)		(18.023)		(23.225)		(23.010)
SCS × Information	−0.003 ***		−0.003 ***		−0.004 ***		−0.004 ***
	(−6.521)		(−6.246)		(−5.616)		(−5.728)
SCS × Governance		−0.441 ***		−0.334 ***		−0.566 ***		−0.536 ***
		(−15.631)		(−14.805)		(−15.590)		(−15.634)
Year	yes	yes	yes	yes	yes	yes	yes	yes
Industry	yes	yes	yes	yes	yes	yes	yes	yes

Note: *** p < 0.01, ** p < 0.05.

reports the heterogeneity test results for cultural synergy. In Columns (1), (3), (5), and (7), the regression coefficients of supply chain structure (SCS) and the interaction term (SCS × Cooperation) are significantly negative at the 1% level. In Columns (2), (4), (6), and (8), the regression coefficients of supply chain structure (SCS) are significantly negative at the 1% level, and the coefficients for the interaction term (SCS × Confucianism) are significantly negative at the 1%, 1%, 5%, and 5% levels, respectively.

These findings suggest that firms with stronger cultural synergy experience a more significant positive impact of supply chain structure diversification on high-quality corporate development.

4.6.4. The Impact of Information Symmetry

In the context of supply chain structure diversification, information transparency acts as a “digital bridge” connecting diverse entities. It enhances collaboration efficiency and risk visibility, driving the transformation toward agile and resilient supply chains. Meanwhile, internal control serves as the “institutional cornerstone” for the healthy operation of supply chains by ensuring compliance and sustainability of diversification strategies through risk management and process optimization. Together, these elements achieve short-term cost optimization and efficiency gains, while laying the foundation for long-term innovation, upgrading, and ecosystem building. They jointly construct a transparent, efficient, and controllable modern supply chain system, providing core competitiveness for high-quality corporate development. Column B of Table 14 reports the heterogeneity test results for information symmetry. In Columns (1), (3), (5), and (7), the regression coefficients of supply chain structure (SCS) and the interaction term (SCS × Information) are significantly negative at the 1% level. In Columns (2), (4), (6), and (8), the regression coefficients of supply chain structure (SCS) and the interaction term (SCS × Governance) are also significantly negative at the 1% level.

These results suggest that firms with higher levels of information symmetry experience a more pronounced positive impact of supply chain structure diversification on high-quality corporate development.

5. Conclusions and Implications

5.1. Research Findings

As a cornerstone of economic development, high-quality enterprise growth plays a vital role in sustaining overall economic performance. In today’s environment of diversified supply chain competition, enhancing collaboration among supply chain partners and building diversified, stable, and secure supply chain structures have become key strategic priorities for economies seeking to modernize their industrial systems. Against this background, this study investigates supply chain structure diversification as a critical pathway to achieving high-quality development. Using econometric methods and empirical analysis, we examine whether and how supply chain diversification promotes firm-level high-quality development. The findings are as follows.

Firstly, supply chain structure diversification significantly promotes high-quality enterprise development by overcoming the resource constraints associated with single-chain dependencies. By integrating resources across entities, regions, and sectors—such as technology, capital, channels, and information—diversification creates a “resource pool” that supports firm growth while mitigating the adverse operational effects of supply chain volatility.

Secondly, diversification provides access to heterogeneous resources and reduces internal path dependence. It introduces both competitive pressure and cooperative incentives, encouraging firms to increase R&D investment and optimize innovation models. Through these mechanisms, diversification enhances the core momentum of high-quality development driven by technological and organizational innovation.

Thirdly, diversification promotes sustainable resource allocation. It lessens reliance on scarce or non-renewable inputs, supports greener and more efficient production, and improves environmental and social performance. The resulting resilience of diversified supply chains reinforces both environmental sustainability and long-term growth quality.

Fourthly, diversification strengthens risk management and cost control. By dispersing operational and financial risks, it compels firms to establish cross-organizational monitoring and coordinated response systems, thereby improving their ability to withstand shocks. A more stable operating environment allows firms to invest in R&D and long-term strategic initiatives, ensuring continuity in high-quality development.

Fifthly, digitalization plays a pivotal moderating role. Digital tools such as the Internet of Things (IoT), blockchain, and big data analytics enhance the efficiency of information transmission and alleviate the “information fragmentation” challenge inherent in diversification. They also improve cross-enterprise resource integration, amplify the mediating effects of diversification on innovation and risk management, and ultimately strengthen its positive impact on high-quality development.

Finally, a comprehensive framework of diversification effects emerges. In the relationship between supply chain structure diversification and high-quality enterprise development, diversification serves as the foundation that provides essential resource reserves; innovation, sustainability, and risk resilience act as mediating bridges that translate those reserves into performance outcomes; and digitalization, operational efficiency, cultural synergy, and information symmetry serve as key boundary conditions that magnify these benefits. Given that our data are drawn from Chinese A-share listed firms within a specific geopolitical and economic context, future extensions could incorporate international samples to validate the external generalizability of these results.

Despite its robust empirical findings, this study faces certain limitations. It primarily examines the causal effects, mechanisms, and boundary conditions of diversification at the focal firm level. Although we note the positive spillovers of digitalization for supply chain ecosystems, our data cannot fully capture the potential externalities—positive or negative—that diversification strategies may impose on upstream and downstream SMEs. Consequently, our conclusions mainly describe the link between focal firm decisions and their performance outcomes. Future research should investigate whether such spillover effects significantly influence the overall coordination, stability, and equity of supply chain ecosystems.

5.2. Policy Implications

Based on the empirical results and theoretical analysis above, several policy and managerial implications can be drawn from the perspectives of enterprise management, industry development, and macroeconomic governance.

Firstly, firms should gradually reduce their dependence on single suppliers or customers by adopting differentiated strategies for different market segments. This can be achieved by exploring non-traditional application scenarios, participating in cross-industry exhibitions, and engaging key opinion leaders (KOLs) for targeted marketing to expand new customer bases. On the supply side, firms should diversify sourcing regions, increase the number of critical suppliers, and maintain a primary-to-backup supplier ratio that keeps reliance on any single supplier below 30%.

Secondly, firms should actively enhance supply chain digitalization to maximize the empowering effects of digital technologies. Implementing digital Supplier Relationship Management (SRM) systems can enable real-time synchronization of orders, inventories, and quality data while setting up automatic “delivery delay alerts.” Establishing Customer Success Management (CSM) teams for key accounts and applying IoT sensors to track logistics data in real time can improve visibility. Integrating these data through big-data analytics and blockchain-based platforms strengthens trust, traceability, and responsiveness throughout the supply chain.

Thirdly, it is essential to foster a corporate culture emphasizing long-term cooperation and mutual benefit. Supplier selection processes should incorporate an assessment of “cultural compatibility,” favoring partners who share Confucian values of integrity and reciprocity. Regular partner summits, joint training programs, and executive exchanges can further reinforce cultural synergy and relational governance.

Fourthly, at the industry level, unified digitalization standards and shared data infrastructures should be promoted. Industry associations, key enterprises, and technology providers should jointly establish standards for data interfaces, cybersecurity, and privacy protection to lower the digital transformation costs faced by SMEs. Industry-wide ESG evaluation systems should also be developed, linking firms’ ESG performance—such as carbon emissions, labor rights, and ethical sourcing—to supply chain participation and order allocation.

Finally, from a policy perspective, governments should strengthen the institutional foundation for supply chain resilience. Differentiated incentive programs can guide firms in adopting digital and green technologies, while a national-level supply chain risk-monitoring platform could provide early-warning signals and periodic vulnerability assessments for key sectors.

In the era of globalization and digital transformation, Chinese enterprises’ supply chain management practices and development models are increasingly becoming aligned with global principles of efficiency, innovation, sustainability, and security. The diversification and digitalization of supply chains not only enhance firms’ competitiveness and resilience but also contribute to sustainable global growth—particularly in labor-intensive industries, emerging markets, and digital-driven sectors.