Supply Chain Relationships, Resilience, and Export Product Quality: Analysis Based on Supply Chain Concentration

Abstract

Supply chain security plays a critical role in ensuring the stable and continuous operation of society. Moreover, enhancing the quality of export products is crucial for improving environmental sustainability, as it helps reduce waste emissions and other related factors. Therefore, this paper employs data from Chinese A-share-listed companies and customs data from 2001 to 2015 to investigate this relationship. The main findings are as follows: (i) The supply chain concentration negatively impacts the quality of export products, a finding that remains robust after testing. (ii) In some firms, such as those where top executives possess digital-related expertise, the adverse effects of the supply chain concentration are likely mitigated. (iii) The channels through which the supply chain concentration affects export product quality may include firm size, productivity, and supply chain efficiency. (iv) Enhancements in infrastructure resilience, firm structure resilience, and industrial structure resilience through investments in regional fixed assets, overseas subsidiaries, and advancements in industrial structure, respectively, are likely to mitigate the negative impacts of the supply chain concentration. These conclusions may hold significant value for promoting both societal and environmental sustainability.

1. Introduction

The emergence of the COVID-19 pandemic has further underscored the importance of supply chain security, making it a significant topic of research in recent years. Countries around the world are striving to devise and implement a series of policy measures to ensure supply chain security or to mitigate the potential negative impacts of supply chain insecurity. However, according to the perspectives of Raman and Shahrur [1], Acemoglu et al. [2], and Kolay et al. [3], a higher supply chain concentration increases the risk faced by firms if major trading partners, whether suppliers or customers, encounter problems. Thus, the supply chain concentration may have a significant impact on a firm’s supply chain security. Supply chain concentration refers to the proportion of transactions with a firm’s top trading partners relative to all its transactions; the higher this proportion, the greater the supply chain concentration [4]. Simultaneously, the quality of export products is also an important area of research. Enhancing export product quality contributes to environmental protection by reducing waste emissions and other related aspects, thereby promoting environmental sustainability [5]. Therefore, the impact of the supply chain concentration on export product quality represents a valuable research question. The following sections will analyze this impact in detail.

Although research on the supply chain concentration has been ongoing for a long time, recent years have witnessed an amplified scholarly focus on this topic, possibly driven by the impact of the COVID-19 pandemic. However, the conclusions of existing literature regarding the effects of the supply chain concentration are not entirely consistent. Some studies argue that the supply chain concentration can be beneficial for firms, while others suggest that it may have negative effects.

Given that technological capability and financial resources are critical factors for firms in producing high-quality products, the literature on how the supply chain concentration impacts these aspects is closely related to this study.

Regarding technological capability, there are differing perspectives in the current literature. Some studies investigate scenarios in which suppliers and customers may influence each other, while others focus on scenarios where such mutual influence does not occur. In situations where suppliers and customers do influence each other, Peters [6] took a novel approach by examining the impact of customer concentration on R&D investment, while also focusing on the concentration level of suppliers. Using data from the German automotive industry, the study found that when the supplier market is less concentrated, a firm’s R&D investment decreases as the customer market concentration increases. Subsequently, analyzing the impact of the supply chain concentration from a theoretical perspective could provide a more systematic approach to the issue, as suggested by Krolikowski and Yuan [7]. Their study, conducted within the framework of incomplete contract theory, resource dependence theory, and transaction cost economics, also emphasized the issue of switching costs. It found that an increase in the customer supply chain concentration may promote supplier innovation; however, when customers possess strong bargaining power, it may inhibit supplier innovation. In scenarios where suppliers and customers do not influence each other, Zhu et al. [8], focusing on inventory management, concluded that an increase in both supplier and customer concentrations can benefit a firm’s productivity. The study also made a notable contribution by identifying the substitutive role of digitalization in mitigating the effects of the supply chain concentration. Jiang et al. [9] found that the supplier concentration negatively affects firm innovation, while the customer concentration has a positive impact. Furthermore, when emphasizing ownership differences, they discovered that the supply chain concentration had a greater impact on the innovation of private firms. Liu et al. [10] focused on comparing the impacts on different types of innovation and found that the customer concentration may lead firms to prioritize incremental innovation over breakthrough innovation.

Regarding financial resources, the current literature also presents mixed views on the impact of the supply chain concentration on a firm’s ability to secure funding. Some studies analyzed this impact within the scenarios of overall financing conditions, while others investigated its effects specifically within the scenarios of bank loans. In the research on overall financing, Upson and Wei [4] emphasized the issue of information asymmetry and innovatively analyzed the income differences for investors facing increased information asymmetry. Their study found that an increase in the supply chain concentration may be more beneficial for firm financing. When firms engage in diversified procurement, partnering with multiple suppliers, their equity and debt costs may rise, and they may face more severe information asymmetry issues. Additionally, firms are more negatively affected when suppliers are in poorer financial condition. However, Fang et al. [11], using data from small and medium-sized firms, found that an increase in the supplier concentration negatively impacts credit ratings, while an increase in customer concentration positively affects credit ratings. Moreover, as different financing channels may be impacted differently, a more detailed analysis is important. Since bank loans are often the most common financing channel, they represent a particularly significant area of research. In banking loan research, Campello and Gao [12] made a noteworthy contribution by examining both pricing and non-pricing aspects in loan contracts. They found that a higher customer concentration leads to higher loan interest rates, more loan restrictions, and shorter loan terms. Furthermore, increased customer concentration also affects the duration and intimacy of a firm’s relationship with banks, which may hinder firm financing. Shi et al. [13] conducted a noteworthy analysis based on signaling theory and found that the customer concentration has a favorable impact on bank loans, while the supplier concentration has an unfavorable impact. Ma and Gao [14], focusing on the role of supply chain relationships in sustainable development, innovatively examined the impact of the supply chain concentration on the scale, cost, and duration of loans from different loan characteristics. Their study discovered that an increased supply chain concentration has a positive impact on all three aspects. Furthermore, when there is less uncertainty among supply chain partners, the positive effects of the supply chain concentration on loan size, cost, and term are even greater. However, the study suggested that a more systematic analysis of supply chain operations is needed, utilizing more sophisticated theoretical models.

In the literature on export product quality, existing research primarily focuses on measuring export product quality and identifying its influencing factors.

Regarding the measurement of export product quality, representative studies include those of Khandelwal et al. [15] and Shi and Shao [16]. Khandelwal et al. [15] constructed an empirical model based on the prices and quantities of exported products to measure export product quality, controlling for important product fixed effects and destination–year fixed effects. In this study, export product quality was measured as the ratio of residuals from the regression to the difference between elasticity and one. In contrast, Shi and Shao [16] measured export product quality by utilizing product trade demand and standardized the quality metrics to facilitate better comparisons. This methodology is widely used in research on China’s export product quality. Both papers share similarities, likely because they adhere to relevant export trade theories.

In addition to measuring export product quality, which is crucial for its improvement, the factors affecting export product quality have received considerable attention. Regarding the factors influencing export product quality, the existing literature can be divided into two categories: macro and micro. At the macrolevel, macroeconomic factors have a broader impact on export product quality, leading many studies to address this issue. Furthermore, due to the direct impact of internationalization factors on export product quality, some literature has examined the influence of internationalization-related factors. Additionally, as some studies argued that the government is a major player in shaping the macroenvironment, the literature has also explored government-related factors.

In the literature on internationalization-related factors, some studies examined scenarios that focus on tangible factors, while others explored scenarios that emphasize intangible factors. Regarding tangible factors, Anwar and Sun [17] made a marginal contribution by employing a firm heterogeneity theoretical model. Their study found that a higher presence of foreign firms within an industry positively affects export quality. Zhang et al. [18] took a relatively unique perspective by examining the impact of internationalization on export quality from the standpoint of foreign-invested banks. They found that foreign-invested banks have a beneficial effect on export quality. Additionally, this study emphasized the distinction between different types of trade, concluding that the positive impact of foreign investment is more pronounced in general trade enterprises. In terms of intangible factors, Xiong et al. [19] made a marginal contribution by examining the impact of foreign divestment. Their analysis revealed that foreign divestment has a significantly negative effect on export quality. Hayakawa et al. [20] focused on the differences in foreign direct investment (FDI) within the service sector, finding that FDI in the service industry promotes improvements in export product quality. However, this effect is more pronounced for foreign-invested enterprises. Oladi and Beladi [21] innovatively studied the “spillover effect” of multinational corporations’ investments in underdeveloped regions on product quality. The research concluded that under conditions of price competition, the “spillover effect” of multinational corporations’ investments in product quality in underdeveloped regions may benefit the improvement of export quality in these areas. Additionally, Yan et al. [22] emphasized the differences in firm productivity and found that outward foreign direct investment (OFDI) positively affects export product quality, with a more significant impact on low-efficiency firms. However, Lu et al. [23] uniquely discovered that an increase in FDI may have a negative impact on export quality. Consequently, among the literature on tangible and intangible factors related to internationalization, it is more often believed that internationalization has a beneficial impact on improving export product quality. Regarding the government-related factors, some studies focused on scenarios examining government revenue and expenditure, while others focused on scenarios where the increasingly serious issue of environmental degradation is a key concern. Regarding government revenues and expenditures, Leng et al. [24] made a marginal contribution by employing dynamic analysis to conclude that fiscal pressure on the government may positively impact export product quality. However, Wu et al. [25] took a novel approach by examining the issue from the perspective of debt acquisition, discovering that an increase in government debt negatively impacts export quality. As environmental degradation has become more severe, environmental protection has garnered greater attention. Given that effective environmental protection largely depends on government regulations, many studies have investigated the impact of these regulations on export product quality. In terms of the impact of environmental regulations, Xiong and Zhu [26] argued that such regulations could potentially improve export product quality. Moreover, the study emphasized the differences between upstream and downstream cities, concluding that the effect of environmental protection on export product quality is stronger for downstream firms. However, He and Tang [27] marginally contributed by analyzing the local governments’ responses to central policies, finding that local environmental policies could negatively impact export product quality. In summary, both government revenues and expenditures, as well as environmental protection, can have either positive or negative effects on export product quality.

At the microlevel, since micro-factors tend to have a more focused impact on export product quality, many studies have been conducted in this area. Since technological capability and capital are important for firms to produce high-quality export products, the literature surrounding these factors is likely to be more pertinent to this study.

Regarding technological capability, given the direct relationship between a firm’s technological level and export product quality, this area has garnered significant scholarly attention. Among the current research frontiers, digitalization and robotic intelligence stand out.

In terms of digitalization, some studies focused on scenarios where the relationship between digitalization and export quality is linear, while others examined scenarios where this relationship is nonlinear. In the context of linear relationships, Chiappini and Gaglio [28] made a marginal contribution by employing an extended gravity model, finding that digitalization positively impacts export quality. Wang and Ye [29] also found that digitalization enhances export quality. However, the study innovatively revealed a negative effect of digitalization, specifically in widening the wage gap within firms. Nonetheless, it primarily focused on firm-level export quality, leaving product-level export quality for further investigation. Moreover, due to data availability constraints, the study only used data from 2007 to 2015. Zhang and Duan [30] emphasized that under the context of market integration, digitalization can improve export quality; however, the negative moderating effect of market segmentation should be taken into account. Since excessive development in many areas can lead to negative outcomes, it is crucial to verify whether overdevelopment in digitalization, despite its advanced nature, might have adverse effects. Therefore, investigating the nonlinear impacts of digitalization could be valuable. In the context of nonlinear relationships, Zhang and Duan [31] found that the effect of digitalization on export quality follows an inverted U-shape. Furthermore, this study placed significant emphasis on the importance of digital infrastructure. It found that insufficient support from infrastructure for digitalization may reduce the beneficial effects of digitalization on export quality, particularly on the left side of the “inverted U-shaped” curve. However, the discussion in this study on the common advantage of enhanced information availability through digitalization could be further elaborated. Wang et al. [32], focusing on electromechanical products, also concluded that the relationship between digitalization and export quality is inverted U-shaped. However, the study reached this conclusion through theoretical analysis alone, indicating that further empirical testing is needed to validate the findings. In summary, in the context of linear relationships, the literature generally views digitalization as beneficial for improving export product quality. In contrast, in nonlinear relationships, the literature suggests that the impact of digitalization on export quality is inverted U-shaped.

In terms of robotics, some studies focused on scenarios where the relationship is linear, while others examined scenarios where the relationship is nonlinear. However, as research in this field has only recently begun to develop, the current body of literature may still be limited, but it is expected to grow in the future. In the context of linear relationships, DeStefano and Timmis [33] studied the relationship between robotics and export product quality, finding that the use of robots improves export product quality, particularly for products that initially had lower quality. This effect is also more pronounced in developing economies, where export product quality tends to be lower. Furthermore, the authors of [33] emphasized how variations in types of robotics affect export product quality across different countries and regions. However, due to data availability, the study only used country–industry-level data, lacking a more granular analysis of different institutions within industries. Lin et al. [34] also found that the application of robotics has a positive impact on export product quality. Moreover, they highlighted that an extended import period enhances the beneficial effects of robotics on export quality. However, while robotics is considered an advanced technological tool, it raises an important question: could excessive use of robotics lead to adverse effects? Therefore, exploring the nonlinear impact of robotics application could be meaningful. In the context of nonlinear relationships, Lu et al. [35] conducted an innovative analysis using an endogenous quality selection model and found that the effect of robotics on export product quality follows an inverted U-shape. The analysis also concluded that robotics impacts export product quality through channels such as improving productivity and suppressing innovation. In summary, the views on robotics application in both linear and nonlinear relationships are likely similar to those in the digitalization literature. In linear relationships, the literature generally suggested that robotics application is beneficial for improving export product quality. In nonlinear relationships, the literature indicated that the impact of robotics on export product quality is inverted U-shaped.

Moreover, since improved technological capabilities require investment to better produce high-quality products, financial resources are also a crucial factor in producing high-quality export products. Therefore, many studies have investigated the impact of firm financing on export product quality. Among these studies, some conducted their analysis within the scenarios of overall corporate financing and export product quality, while others focused on more specific scenarios, conducting their research within the scenarios of bank funding and export product quality. Regarding overall financing, Hu et al. [36] found that financing constraints negatively affect export product quality. Moreover, this study, within the context of advancing market integration in China, emphasized the importance of improving the market system. It found that enhancements in the market system can mitigate the adverse effects of financing constraints on export product quality. Choi [37] innovatively employed a heterogeneous firm model, incorporating credit constraints and quality selection. The study found that a reduction in financing costs benefits the improvement of a firm’s export product quality. Ma et al. [38] found that green finance benefits the improvement of export product quality. Additionally, this study made a marginal contribution by examining the impact of green insurance, finding that green insurance positively influences the improvement of export product quality. Given that different financing methods can have varying impacts, and since bank loans may be the most common method for firms to obtain financing, many studies have explored the effects of bank loans on export product quality. In terms of bank financing, Sui et al. [39] argued that bank loans positively affect export product quality. Furthermore, this study highlighted that state-owned enterprises may experience greater benefits from bank loans. From a reverse perspective, Qiu et al. [40] examined bank loan restrictions and found that easing these restrictions can enhance export product quality. Additionally, the study also highlighted the differences in financing dependence among firms, finding that firms with higher levels of financing dependence experience more significant impacts from the relaxation of bank loan restrictions. Accordingly, the literature generally suggested that improvements in financing are beneficial for enhancing export product quality.

Based on the above, there may not yet be a unified conclusion regarding the impact of the supply chain concentration, and there is limited literature on how it affects the quality of exported products. Specifically, the impact of the supply chain concentration on technological capability, which is potentially the most critical factor for firms producing high-quality export products, remains inconclusive.

Consequently, this study aimed to investigate whether the supply chain concentration affects export product quality, thereby providing insights to enhance supply chain management practices and improve the quality of exported products. By analyzing data from Chinese A-share (A-shares refer to common stocks issued by companies within mainland China, which are subscribed to and traded in RMB by domestic institutions, organizations, or individuals) listed companies on the Shanghai and Shenzhen stock exchanges between 2001 and 2015, this research established that an increased supply chain concentration adversely affects export product quality, a conclusion supported by robustness checks. Furthermore, heterogeneity analysis revealed that the adverse effects of the supply chain concentration are less pronounced in technology-intensive firms, firms with banking relationships, firms with CEOs having overseas backgrounds, and firms with top executives possessing expertise in digital-related fields. Mechanism analysis indicated that the supply chain concentration may influence export product quality through mechanisms including firm size, production efficiency, and supply chain effectiveness. Lastly, moderation effect analysis suggested that regional fixed asset investment, overseas subsidiaries, and advanced industrial structures can positively moderate the negative impacts of the supply chain concentration on infrastructure resilience, firm structure resilience, and industry structure resilience.

This paper may offer the following marginal contributions: (1) Providing literature support. Given the limited research on the impact of the supply chain concentration on export product quality, this study aimed to fill that gap. Moreover, as the COVID-19 pandemic has underscored the critical importance of supply chain issues, pursuing high product quality is essential for the stable development of international trade and supports environmental protection. Therefore, studying the relationship between these two crucial aspects is of significant practical relevance. (2) Conducting targeted heterogeneity analysis. This includes examining two crucial factors that influence the production of high-quality export products—technological capability and financing—as well as two key aspects that affect the impact of supply chain concentration: the CEOs’ overseas background and the top executives’ expertise in digital-related fields. Since technology, capital, and management are closely linked to the subject of this study, relevant research may offer important practical insights for firms with differences in these areas to better manage the relationship between supply chain concentration and export product quality. Therefore, this analysis is likely to be of significant practical relevance. (3) Conducting a comprehensive analysis of impact pathways. This study examined the influence of supply chain concentration from the dual perspectives of size and efficiency, both of which are critical in firms’ operations. The analysis revealed that firm size, production efficiency, and supply chain efficiency are pathways through which the supply chain concentration impacts export product quality. Moreover, analyzing both scale and efficiency together may better improve export product quality. On one hand, increasing efficiency while expanding scale could lead to a more effective improvement in export product quality. On the other hand, increasing scale while improving efficiency might provide a stronger material foundation for enhancing product quality. Therefore, simultaneously analyzing both scale and efficiency could have significant practical implications for systematically mitigating the negative effects of the supply chain concentration on export products. (4) Analyzing the effects of resilience. This paper examined the role of resilience from the perspectives of infrastructure resilience, firm structure resilience, and industrial structure resilience. The findings suggested that stronger resilience in these areas can mitigate the negative impact of the supply chain concentration on export product quality. Therefore, the findings from this part of the study may hold significant practical relevance for understanding how to use external resilience to mitigate the negative impact of the supply chain concentration on export product quality.

The remainder of this paper is structured as follows: Section 2 presents the theoretical analysis and hypotheses, Section 3 explains the research design, Section 4 discusses the empirical results, Section 5 offers further discussion, and Section 6 concludes with policy implications.

2. Theoretical Analysis and Hypothesis

The reasons for an increased supply chain concentration include technological barriers in raw material supply, oligopolistic dominance in supplier markets, and exclusion from foreign markets, forcing firms to transact with a limited number of trading partners to acquire raw materials or sell products. However, the increased concentration of supply chains may affect export product quality through its impact on firm size and efficiency, as detailed below.

2.1. Size

An increased supply chain concentration may negatively impact firm size, thereby adversely affecting the quality of export products. On one hand, in terms of resource crowding-out, a high supply chain concentration allows larger trading partners to exert greater dominance [41]. As a result, firms may be required to reserve resources specifically to ensure timely coordination with these major partners [42,43,44]. For instance, major trading partners may require firms to designate a portion of their cash reserves as dedicated funds exclusively for operations related to these partners. Consequently, a high supply chain concentration can crowd out a portion of the firm’s resources, negatively affecting its size. On the other hand, in terms of operational constraints, when the supply chain concentration is high, firms become more dependent on large trading partners for their operations [3]. This increased dependence can restrict the overall scale of the firm’s operations, thereby reducing the firm size. However, a reduction in firm size can negatively impact the quality of export products. Regarding cost inputs, producing high-quality export products often requires significant cost investments, including machinery and equipment assembly, raw material procurement, and labor hiring [45,46]. Nevertheless, as the firm size decreases, it becomes more challenging for firms to bear the high costs associated with producing high-quality products, possibly forcing them to produce lower-quality export products. Backward induction may have been applied here because firms tend to anticipate potential future scenarios before taking action. Backward induction refers to the process of predicting a firm’s current actions based on the possible outcomes it may face after taking various actions. Thus, using backward induction not only enables a more systematic analysis of the impact on firms but also helps mitigate disruptions in the model caused by the absence of future scenario considerations. With respect to the operating cycle, high-quality export products typically have longer production cycles and may require extended time for export trade. Therefore, firms producing high-quality export products need a robust financial chain capable of withstanding long return cycles. However, smaller firms are less able to endure financial strain, leading them to prefer producing lower-quality export products. Accordingly, this study proposes the following hypotheses:

Hypothesis 1.

An increased supply chain concentration will reduce the quality of export products.

Hypothesis 2.

An increased supply chain concentration will reduce the quality of export products by reducing the firm size.

2.2. Efficiency

An increased supply chain concentration may negatively impact firm efficiency, thereby adversely affecting the quality of export products. In terms of supply chain cooperation, as the supply chain concentration increases, firms become more dependent on major trading partners, making their production activities increasingly reliant on the coordination and cooperation of these key partners. This heightened reliance can restrict the firm’s operations [41], limiting its ability to carry out various activities freely, thereby reducing efficiency. Furthermore, when major trading partners hold greater dominance, they may become less motivated to actively promote or facilitate supply chain collaboration, which could further reduce overall efficiency. In terms of supply chain disruptions, a higher level of supply chain concentration means that if a major trading partner experiences any issues or deviations, the firm, being heavily reliant on that partner, may suffer significant disruptions [1,3], thereby impacting operational efficiency. However, reduced efficiency can lead to a decline in the quality of export products. According to traditional economic theory, decreased efficiency results in slower operations, reducing the output per unit. Consequently, the number of components used in products may decrease, thereby negatively affecting the quality of export products. Based on the above discussion, the following hypothesis is proposed:

Hypothesis 3.

An increased supply chain concentration will reduce the quality of export products by decreasing firm efficiency.

3. Research Design

3.1. Model Setup

To investigate the impact of the supply chain concentration on the quality of export products, we constructed the following model:

$${\mathit{Quality}}_{\mathit{ihdt}}={\alpha }_0+{\alpha }_1{\mathit{SCC}}_{\mathit{it}}+{\alpha }_2{\mathit{Control}}_{\mathit{it}}+{\mathit{Firm}}_i+{\mathit{Product}}_h+{\mathit{Destination}}_d+{\mathit{Year}}_t+{\theta }_{\mathit{ihdt}},$$

(1)

where ${\mathit{Quality}}_{\mathit{ihdt}}$ represents the dependent variable, which is the quality of products $h$ exported by firm $i$ in year $t$ to destination $d$. ${\mathit{SCC}}_{\mathit{it}}$ denotes the supply chain concentration, and ${\mathit{Control}}_{\mathit{it}}$ represent control variables. ${\mathit{Firm}}_i$, ${\mathit{Product}}_h$, ${\mathit{Destination}}_d$, and ${\mathit{Year}}_t$ denote firm fixed effects, product fixed effects, destination fixed effects, and time fixed effects, respectively. ${\theta }_{\mathit{ihdt}}$ stands for the random error term.

3.2. Variable Measurement

3.2.1. Dependent Variable

The dependent variable in this study was the quality of export products, calculated following the methods of Khandelwal et al. [15] and Shi and Shao [16]. The calculation steps are outlined below. The Reference Price Effect may have been considered here. According to traditional economic theory, the Reference Price Effect refers to the phenomenon where buyers are more sensitive to a product’s price when the price is higher relative to other substitute products, and less sensitive when the price is lower. Therefore, the considerations related to the Reference Price Effect in this study were likely aimed at making the research more relevant and reflective of real-world conditions.

Step 1. We constructed the regression model:

$${{\sigma p}_{\mathit{ihdt}}+q}_{\mathit{ihdt}}{=\mathit{Destination}\_\mathit{Year}}_{\mathit{dt}}+{\mathit{Product}}_h+{\epsilon }_{\mathit{ihdt}},$$

(2)

where $p_{\mathit{ihdt}}$ is the natural logarithm of the product price. We included the price variable when calculating export product quality because price is a crucial factor in export products, making it a critical factor in calculating export product quality. Additionally, the selection of other variables followed similar reasoning and will not be further elaborated. $q_{\mathit{ihdt}}$ is the natural logarithm of the product export quantity, ${\mathit{Destination}\_\mathit{Year}}_{\mathit{dt}}$ denotes destination–year fixed effects, and ${\mathit{Product}}_h$ represents product fixed effects.

Step 2. We calculated the export product quality:

$${\mathit{EPQ}}_{\mathit{ihdt}}={\widehat{\epsilon }}_{\mathit{ihdt}}/\left(\sigma -1\right),$$

(3)

where elasticity ($\sigma$) was derived following the method of Fan et al. [47].

Step 3. We standardized the export product quality following Shi and Shao [16]:

$${\mathit{Quality}}_{\mathit{ihdt}}=\left({\mathit{EPQ}}_{\mathit{ihdt}}-{\mathit{EPQ}\_\mathit{min}}_{\mathit{ihdt}}\right)/\left({\mathit{EPQ}\_\mathit{max}}_{\mathit{ihdt}}-{\mathit{EPQ}\_\mathit{min}}_{\mathit{ihdt}}\right),$$

(4)

where ${\mathit{EPQ}\_\mathit{max}}_{\mathit{ihdt}}$ represents the maximum export product quality and ${\mathit{EPQ}\_\mathit{min}}_{\mathit{ihd}}$ represents the minimum. The significance of considering the “Reference Price Effect” in the current context is as follows: Firstly, in the context of high globalization, economic connections between global economies are closely intertwined. Secondly, major global events, such as the Russia–Ukraine conflict, have heightened global instability. Thirdly, the supply chain concentration, as a form of interconnectedness between different entities, is significantly affected by global instability. Lastly, export product quality, as part of international trade, is also heavily influenced by global instability. Therefore, in a highly interconnected yet unstable environment, considerations related to the “Reference Price Effect” may hold great importance.

3.2.2. Independent Variable

The independent variable was the supply chain concentration ($\mathrm{SCC}$), with robustness tests dividing it into supplier concentration ($\mathrm{SC}$) and customer concentration ($\mathrm{CC}$). Following the methods of Krolikowski and Yuan [7] and Tana and Chai [48], we first calculated the proportion of total purchases from the top-five suppliers to the total purchasing amount, and separately. We then separately calculated the proportion of total sales to the top-five customers relative to the total sales amount for each firm. Finally, we then used the average of these two proportions to measure the $\mathrm{SCC}$.

3.2.3. Control Variables

Following the studies by Ciani and Bartoli [42] and DeStefano [33], we selected the following control variables: firm age ($\mathrm{Age}$), calculated by subtracting the year of establishment from the current year, adding one, and then taking the natural logarithm; state ownership ($\mathrm{S}\mathrm{O}$), a dummy variable where 1 indicates the firm is state-owned and 0 otherwise; capital intensity ($\mathrm{Capital}$), measured by the ratio of fixed assets to total assets; ownership concentration ($\mathrm{O}\mathrm{C}$), measured by the shareholding percentage of the largest shareholder; financing constraints ($\mathrm{F}\mathrm{C}$), measured using the WW index. Since the WW index is inversely related to financing constraints, a higher WW index indicates fewer financing constraints faced by the firm.

Reasons for selecting the control variables were as follows. Firm age ($\mathrm{Age}$): Firm age was included because, as a firm ages, its ability to embrace new developments may diminish, making it harder to keep pace with evolving trends. This could result in a relative decline in export product quality. Therefore, to minimize the interference of firm age on the empirical results, firm age was included as a control variable. Additionally, since export product quality in this study was standardized, reflecting a relative level, the empirical results may show a negative value. State ownership ($\mathrm{S}\mathrm{O}$): When firms are state-owned, they often receive greater support from the government. Moreover, state-owned firms are more likely to follow government policies in production, and China has policies aimed at promoting high-quality trade development. As a result, state-owned firms may put more effort into improving export product quality. To minimize interference in the results, state ownership was included as a control variable in the model. However, the empirical result for state ownership was not significant. This is possibly because non-state firms, facing intense international competition, may have realized that improving export product quality is key to growth, leading them to also enhance product quality. This could explain the insignificant empirical result for state ownership. Capital intensity ($\mathrm{Capital}$): Firms with higher capital intensity have more resources to invest in machinery and equipment, which can contribute to improving export product quality. To reduce potential interference in the results, capital intensity was included as a control variable. The negative empirical result was also consistent with expectations. Ownership concentration ($\mathrm{O}\mathrm{C}$): As ownership concentration increases, large shareholders gain more power, which may lead to more self-serving behaviors that negatively impact the firm’s operations, ultimately harming export product quality. Therefore, ownership concentration was included as a control variable to reduce interference in the results. Additionally, the negative coefficient for ownership concentration was reasonable. Financing constraints ($\mathrm{F}\mathrm{C}$): When financing constraints are lower, firms have greater access to capital, which can facilitate improvements in export product quality. To minimize interference in the results, financing constraints were added as a control variable. It is also worth noting that the WW index had a negative relationship with financing constraints, so the negative empirical result was logical.

3.3. Data Source and Processing

This study utilized two main datasets: data from companies listed on the Shanghai and Shenzhen Stock Exchanges in China and customs data. The data from companies listed on the Shanghai and Shenzhen Stock Exchanges in China provided a comprehensive range of information on various aspects of these firms, offering robust support for conducting research. Customs data, compiled by China Customs, included detailed information on all imports and exports, such as types, quantities, prices, codes, and destinations, making it possibly the most widely used dataset for researching issues related to international trade involving China. In the overall dataset, since the earliest available data on supply chain concentration were from 2001, and the calculation of export product quality using Chinese customs data was feasible only up to 2015, we used data from Chinese A-share-listed companies on the Shanghai and Shenzhen Stock Exchanges, along with customs data, for the period from 2001 to 2015. Additionally, following the methodology of Krolikowski and Yuan [7] and Bernini et al. [45], the data underwent the following processing steps: (1) exclusion of listed companies identified as ST, ST*, and PT, (2) retention of only non-financial companies engaged in general trade, and (3) removal of missing and outlier values in key variables. The descriptive statistics of the variables are shown in

Table 1. Descriptive statistics.

Variables	Obs.	Mean	S.D.	Min.	Max.
Quality	397,403	0.5291	0.2626	0	1
SCC	397,403	0.2313	0.1304	0.0250	0.9560
Age	397,403	2.5732	0.4073	0.6931	3.8918
SO	397,403	0.4945	0.5000	0	1
Capital	397,403	0.2066	0.1283	0.0002	0.8000
OC	397,403	0.3881	0.1540	0.0431	0.8855
FC	397,403	−1.0125	0.0721	−1.5050	1.4604

.

4. Empirical Results and Discussion

4.1. Baseline Model Results and Discussion

Table 2. Baseline regression results.

Variables	Dependent Variable: Quality
	(1)	(2)	(3)
SCC	−0.0407 ***	−0.0409 ***	−0.0456 ***
	(−4.9422)	(−4.9316)	(−5.4571)
Age			−0.0311 ***
			(−4.6878)
SO			0.0029
			(0.7618)
Capital			0.0145 **
			(1.9661)
OC			−0.0457 ***
			(−3.7439)
FC			0.0455 ***
			(2.9073)
Constant	0.5386 ***	0.5387 ***	0.6792 ***
	(275.3483)	(273.1744)	(26.8392)
Firm FE	Yes	Yes	Yes
Product FE	Yes	Yes	Yes
Destination FE	Yes	Yes	Yes
Year FE	No	Yes	Yes
N	397,021	397,021	397,021

Note: *** and ** indicate significance at the 1% and 5% levels, respectively. The values in parentheses are t-values.

presents the results of the baseline model. In column (1), the independent variable, supply chain concentration, and the dependent variable, export product quality, are included, with controls for firm, product, and destination fixed effects. Column (2) further incorporates time fixed effects, and finally, column (3) incorporates all control variables. The results showed that in columns (1) to (3) of Table 2, the coefficients for supply chain concentration were all significantly negative at the 1% level. This indicated that an increase in supply chain concentration had a negative impact on the quality of export products, thus supporting Hypothesis 1. Specifically, according to the results in column (3), a one-unit increase in supply chain concentration may reduce the quality of export products by 0.0456 units.

4.2. Robustness Tests

• Endogeneity test: Because firms with a higher export product quality are likely to be stronger and have more trading partners to choose from, they may try to reduce the transaction volume with any single partner to avoid constraints. Therefore, the baseline model may have endogeneity issues. To test for endogeneity, this study used the average supply chain concentration at the provincial and industry levels as an instrumental variable. The results in

  Table 3. Instrumental variable regression results.

  Variables	First Stage	Second Stage
  	Dependent Variable: SCC	Dependent Variable: Quality
  	(1)	(2)
  IV	0.4977 ***
  	(143.1366)
  SCC		−0.0377 **
  		(−1.9646)
  Age	−0.0507 ***	−0.0307 ***
  	(−30.6147)	(−4.5847)
  SO	−0.0063 ***	0.0029
  	(−7.9847)	(0.7767)
  Capital	−0.0096 ***	0.0143 *
  	(−4.9267)	(1.9424)
  OC	0.0305 ***	−0.0461 ***
  	(7.8340)	(−3.7706)
  FC	0.1808 ***	0.0435 ***
  	(22.7838)	(2.6805)
  Firm FE	Yes	Yes
  Product FE	Yes	Yes
  Destination FE	Yes	Yes
  Year FE	Yes	Yes
  Under-identification Test	7408.82 ***
  Cragg–Donald Wald F Statistic	88,846.17
  Kleibergen–Paap Wald RK F Statistic	20,488.08
  N	397,021	397,021

  Note: ***, **, and * denote significance levels of 1%, 5%, and 10%, respectively. The t-value is presented within parentheses.

  show that the instrumental variable passed the under-identification test and the weak instrumental variable test, suggesting that the choice of instrumental variable was reasonable. Moreover, even when using the instrumental variable, the result for the supply chain concentration remained significantly negative. This indicated that the previous estimates were likely not significantly affected by endogeneity.
• Changes in variable measurement: As previously mentioned, the supply chain concentration ($\mathrm{SCC}$) was divided into supplier concentration ($\mathrm{SC}$) and customer concentration ($\mathrm{CC}$) for testing. Referring to the methods of Krolikowski and Yuan [7] and Tana and Chai [48], the supplier concentration ($\mathrm{SC}$) was measured by the proportion of purchases from the top-five suppliers to the total procurement. Similarly, the customer concentration ($\mathrm{CC}$) was measured by the proportion of sales to the top-five customers to the total sales. The results in columns (1) and (2) of

  Table 4. Other robustness tests.

  Variables	Dependent Variable: Quality
  	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)
  SCC			−0.0456 ***	−0.0451 ***	−0.0489 ***	−0.0332 ***	−0.0366 ***	−0.0455 ***	−0.0642 ***
  			(−5.4571)	(−5.3958)	(−5.8771)	(−3.0206)	(−3.7698)	(−5.2829)	(−6.6342)
  SC	−0.0147 **
  	(−2.4380)
  CC		−0.0425 ***
  		(−6.2796)
  Age	−0.0294 ***	−0.0311 ***	−0.0285 ***	−0.0285 ***	−0.0288 ***	−0.0175 *	−0.0401 ***	−0.0253 ***	−0.0290 ***
  	(−4.4443)	(−4.6995)	(−4.2773)	(−4.2773)	(−4.4049)	(−1.6523)	(−5.5305)	(−3.1670)	(−3.3466)
  SO	0.0034	0.0020	0.0037	0.0037	0.0036	0.0180 ***	0.0053	0.0026	−0.0002
  	(0.9019)	(0.5323)	(0.9806)	(0.9806)	(0.9512)	(3.2227)	(1.1920)	(0.6744)	(−0.0397)
  Capital	0.0133 *	0.0160 **	0.0134 *	0.0134 *	0.0085	0.0079	0.0127	0.0126 **	0.0010
  	(1.8076)	(2.1769)	(1.8220)	(1.8220)	(1.1491)	(0.8363)	(1.5201)	(1.6906)	(0.1195)
  OC	−0.0468 ***	−0.0476 ***	−0.0438 ***	−0.0438 ***	−0.0277 **	−0.0364 **	−0.0549 ***	−0.0449 ***	−0.0421 ***
  	(−3.8319)	(−3.8975)	(−3.5768)	(−3.5768)	(−2.2768)	(−2.2057)	(−4.1393)	(−3.6612)	(−3.2645)
  FC	0.0375 **	0.0451 ***	0.0399 **	0.0399 **	0.0524 ***	0.0182	0.0342 *	0.0476 ***	0.0825 ***
  	(2.3993)	(2.8879)	(2.5365)	(2.5365)	(3.3349)	(0.9391)	(1.9280)	(2.7767)	(4.4127)
  PCGDP				−0.0049
  				(−0.3468)
  IA				0.0331 ***
  				(3.4442)
  Constant	0.6600 ***	0.6788 ***	0.6792 ***	0.3213 ***	0.6748 ***	0.6011 ***	0.6834 ***	0.6666 ***	0.7162 ***
  	(26.3422)	(27.0130)	(26.8392)	(3.1265)	(26.7806)	(16.4612)	(24.5059)	(22.9410)	(22.5098)
  Firm FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
  Product FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
  Destination FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
  Year FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
  N	397,021	397,021	397,021	396,988	397,318	240,581	289,388	397,021	363,725

  Note: Columns (1) and (2) present the results when the independent variables were changed to supplier concentration and customer concentration. Column (3) shows the results of altering the export product quality elasticity. Column (4) presents the results after adding control variables. Column (5) displays the results of changing the fixed effects. Column (6) shows the results when only samples from post-2008 were retained. Column (7) presents the results using only manufacturing industry data. Columns (8) and (9) display the results after winsorization and truncation, respectively. ***, **, and * denote significance levels of 1%, 5%, and 10%, respectively. The t-value is presented within parentheses.

  showed that both the supplier concentration and customer concentration had significantly negative impacts on the quality of export products. This indicated that an increase in either the supplier concentration or customer concentration would negatively affect export product quality. Additionally, since there are differences in the elasticity used to calculate export product quality in various studies, we also tested the robustness of the results by changing the elasticity. Following the suggestion of Anderson and Wincoop [49] that the elasticity can be set to 10, which has been adopted in some literature, the export product quality was recalculated with the adjusted elasticity, and the regression analysis was conducted again. The result in column (3) of Table 4 showed that even after changing the elasticity, the impact of the supply chain concentration remained significantly negative. This suggested that differences in elasticity may not fundamentally affect the results.
• Addition of control variables: To enhance the robustness of the previous estimates, additional macrolevel control variables were included in the analysis. Since per capita GDP ($\mathrm{PCGDP}$) and industrial agglomeration ($\mathrm{I}\mathrm{A}$) may affect the results of the baseline model at the macrolevel, these two variables were added to the baseline model for estimation. Per capita GDP ($\mathrm{PCGDP}$) was calculated as the ratio of provincial GDP to the total population, while industrial agglomeration ($\mathrm{I}\mathrm{A}$) was measured using the method of Kosfeld and Titze [50]: measured by the ratio of each province’s share of national industrial added value to its share of national total output. The results in column (4) of Table 4 showed that, even after adding these control variables, the impact of the supply chain concentration on export product quality remained significantly negative, suggesting that the previous estimates were likely robust.
• Adjustment of estimation method: As some literature, such as Brandt et al. [51], analyzed import and export issues at the four-digit code level, the product fixed effects were also tested at the four-digit code level. The results in column (5) of Table 4 showed that, even after changing the fixed level of product fixed effects, the impact of the supply chain concentration remained significantly negative. This indicated that different levels of fixed effects may not fundamentally affect the results.
• Replacement of samples: Since different samples may yield different results, several methods were used to replace the samples for testing. Firstly, considering that the global financial crisis of 2008 may have impacted import and export trade [52], only samples from post-2008 were retained for regression, with the results presented in column (6) of Table 4. Secondly, as the export situation of non-manufacturing products may differ [51], only manufacturing data were retained for re-estimation, with the results shown in column (7) of Table 4. Finally, as to whether to applying winsorization and truncation may lead to different results, these methods were tested, and the results are presented in columns (8) and (9) of Table 4, respectively. The results in columns (6) to (9) of Table 4 indicated that the impact of the supply chain concentration remained significantly negative, suggesting that differences in samples may not significantly affect the results.

5. Further Discussion

5.1. Heterogeneity Analysis

The analysis of heterogeneity separately integrated the themes of export product quality and supply chain concentration. On one hand, as technology and capital may be critical factors for firms producing high-quality products, the analysis was conducted based on whether firms belong to the high-tech industry and whether they have banking connections. On the other hand, as an increased supply chain concentration will lead to varying impacts depending on management capability, the analysis focused on whether the CEO had an overseas background and whether top executives had expertise in digital-related fields. Since the heterogeneity analysis was conducted in a problem-oriented manner, we first focused on analyzing heterogeneity in relation to export product quality.

5.1.1. High-Tech Industry

Since technological factors may alter the impact of the supply chain concentration on export product quality, the analysis categorized firms based on whether they belong to the high-tech industry. The results in columns (1) and (2) of

Table 5. Heterogeneity analysis.

Variables	High-Tech Industry		Connection with Banks		Overseas Background		Digital-Related Expertise
	Yes	No	Yes	No	Yes	No	Yes	No
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
SCC	−0.0154	−0.0639 ***	−0.0254	−0.0497 ***	0.0491	−0.0387 ***	−0.0296	−0.0478 ***
	(−1.3375)	(−4.7753)	(−0.9407)	(−5.4699)	(1.1900)	(−3.5572)	(−0.6332)	(−5.5043)
Age	−0.0430 ***	0.0022	−0.1288 ***	−0.0222 ***	0.0117	−0.0473 ***	−0.0697 **	−0.0248 ***
	(−5.5512)	(0.1595)	(−3.5844)	(−3.1901)	(0.1868)	(−5.6299)	(−2.4677)	(−3.4963)
SO	0.0094 *	−0.0027	0.0366	−0.0005	0.0295 **	0.0135 **	0.0673 ***	−0.0051
	(1.8486)	(−0.4574)	(1.3580)	(−0.1229)	(2.1175)	(2.1464)	(8.0986)	(−1.2321)
Capital	−0.0032	0.0455 ***	0.0074	0.0166 **	−0.1624 ***	0.0299 ***	−0.0252	0.0211 ***
	(−0.3169)	(3.9029)	(0.2680)	(2.0455)	(−5.8984)	(3.0373)	(−0.6984)	(2.6899)
OC	−0.0188	−0.0378 *	0.2012 ***	−0.0441 ***	−0.3781 ***	−0.0313 *	0.0409	−0.0569 ***
	(−1.1955)	(−1.7850)	(2.7043)	(−3.4139)	(−3.4916)	(−1.9053)	(0.6485)	(−4.5300)
FC	0.0389 *	0.0163	0.0258	0.0477 ***	0.3774 ***	0.0010	−0.0176	0.0529 ***
	(1.9273)	(0.6179)	(0.5061)	(2.8549)	(6.1175)	(0.0496)	(−0.2564)	(3.2451)
Constant	0.6707 ***	0.5869 ***	0.8314 ***	0.6554 ***	1.0310 ***	0.6627 ***	0.6419 ***	0.6799 ***
	(21.6714)	(12.5898)	(7.0582)	(24.3601)	(5.1087)	(20.9261)	(6.1995)	(25.3979)
Firm FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Product FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Destination FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	223,851	172,477	59,745	336,771	24,542	233,066	30,772	363,676

Note: ***, **, and * denote significance levels of 1%, 5%, and 10%, respectively. The t-value is presented within parentheses.

indicated that for non-high-tech firms, the coefficient for supply chain concentration remained significantly negative. However, for high-tech firms, the impact of the supply chain concentration was not significant. This suggested that the negative impact of an increased supply chain concentration was smaller for high-tech firms. This may be because these firms can mitigate the adverse effects of the supply chain concentration through their technological capabilities and external knowledge spillovers. Internally, high-tech firms are likely to possess higher technological levels. As the supply chain concentration increases, these firms can leverage advanced technology to neutralize the adverse effects, thereby mitigating the decline in export product quality. Externally, high-tech firms are more likely to engage in technology exchanges and participate in technology conferences. These activities enable them to benefit from external knowledge spillovers, which helps mitigate the negative impact of the supply chain concentration.

5.1.2. Connection with Banks

Since financing constraints are a primary reason for the impact of the supply chain concentration on export product quality, the nature of their banking relationships may influence the impact of the concentration, thereby warranting further investigation. Therefore, firms are categorized based on their connection with banks for regression analysis, as shown in columns (3) and (4) of Table 5. Drawing on Kupiec et al. [53], the classification of whether firms are connected with banks is based on whether firms own bank shares. When firms hold bank shares, they may engage in more cooperative exchanges with banks. Moreover, holding bank shares inherently indicates a closer relationship with banks, as firms would be less inclined to invest in bank shares without an existing strong connection.

The results indicated that when firms did not have connections with banks, the coefficient for supply chain concentration remained significantly negative. However, when firms had connections with banks, the effect of the supply chain concentration was not significant. This may be because firms with bank connections can better access financial support from banks. Additionally, bank loans are often provided in cash, which is more liquid compared to other forms of external capital that may involve delays. Consequently, firms with bank connections not only receive more loan assistance but also gain access to more liquid resources, which helps mitigate the negative effects of an increased supply chain concentration. Thus, firms with bank connections may experience less negative impact from an increased supply chain concentration.

5.1.3. CEO’s Overseas Background

Since CEOs with overseas backgrounds may have been exposed to more advanced management practices, they might possess stronger management capabilities. Therefore, it is worth examining whether the impact of the supply chain concentration on export product quality differs when the CEO has an overseas background. When analyzing CEO management heterogeneity, the selection of indicators was closely aligned with the research theme, which may more effectively contribute to improvements in relevant areas. The reason for selecting the CEO’s overseas background as a research variable was that when a CEO has an overseas background, their enhanced management capabilities are more focused on import–export trade, which is more relevant to the topic of export product quality than general management indicators. Therefore, the analysis of the CEO’s overseas background may be more beneficial for firms in mitigating the negative impact of the supply chain concentration on export product quality. The results in columns (5) and (6) of Table 5 indicated that when the CEO did not have an overseas background, an increase in the supply chain concentration had a significantly negative impact on export product quality. However, when the CEO had an overseas background, the impact of the increased supply chain concentration on export product quality was not significant. This may be because CEOs with overseas backgrounds have a better understanding of which issues to address and which management measures to implement when the supply chain concentration increases, thereby mitigating its negative effects. Consequently, the adverse impact of an increased supply chain concentration on export product quality was smaller when the CEO had an overseas background.

5.1.4. Top Executives’ Digital-Related Expertise

The estimation results regarding whether top executives graduated from digital-related majors are presented in columns (7) and (8) of Table 5. Specifically, for firms whose top executives did not major in digital-related fields, the supply chain concentration had a significantly negative effect on export product quality. In contrast, for firms with top executives who majored in digital-related fields, the effect of the supply chain concentration was not significant. This difference in impact may be due to the ability of top executives with a background in digital-related majors to better leverage digital technologies, allowing them to more quickly and efficiently gather information. As a result, when the supply chain concentration increases, these firms can obtain critical information faster, mitigating the negative effects of an increased supply chain concentration. Additionally, because top executives with digital-related expertise can promptly track changes among key suppliers and customers, their firms may reduce the resources reserved for timely coordination with large trading partners, alleviating resource constraints. Consequently, in firms where top executives graduated from digital-related majors, the adverse impact of an increased supply chain concentration on export product quality is likely to be less pronounced.

5.2. Mechanism Analysis

To analyze the channels through which the supply chain concentration affects export product quality, the following model was constructed:

$$M_{\mathit{it}}={\beta }_0+{\beta }_1{\mathit{SCC}}_{\mathit{it}}+{\beta }_2{\mathit{Control}}_{\mathit{it}}+{\mathit{Firm}}_i+{\mathit{Year}}_t{+ϵ}_{\mathit{it}},$$

(5)

where $M_{\mathit{it}}$ represents the mechanism variables, including firm size (${\mathit{Size}}_{\mathit{it}}$), total factor productivity (${\mathit{TFP}}_{\mathit{it}}$), and supply chain efficiency (${\mathit{SCE}}_{\mathit{it}}$). Other aspects of the model remained consistent with the baseline model discussed earlier. Firm size ($\mathit{Size}$) was measured using both the firm’s asset size ($\mathit{Size}1$) and the number of employees ($\mathit{Size}2$), with the natural logarithm applied to each variable. Total factor productivity ($\mathit{TFP}$) was assessed using the fixed effects methodology, while supply chain efficiency ($\mathit{SCE}$) was measured as the natural logarithm of the ratio of 365 days to the inventory turnover period, following Danese and Bortolotti [54]. The inclusion of total factor productivity ($\mathit{TFP}$) and supply chain efficiency ($\mathit{SCE}$) was because a firm’s efficiency encompasses both production and distribution aspects: total factor productivity focuses on production efficiency, while supply chain efficiency emphasizes distribution efficiency.

The results in

Table 6. Mechanism analysis.

Variables	Size 1	Size 2	TFP	SCE
	(1)	(2)	(3)	(4)
SCC	−0.7836 ***	−1.1370 ***	−0.4269 ***	−0.5728 ***
	(−6.0544)	(−7.1508)	(−2.7758)	(−5.0661)
Age	0.3501 ***	0.1809	0.4987 ***	−0.0619
	(2.7791)	(1.2717)	(3.1295)	(−0.6085)
SO	0.1177 **	0.0715	0.0919	0.0755 *
	(2.0863)	(1.0490)	(1.5956)	(1.9113)
Capital	−0.4160 ***	0.2640 *	−0.2498 **	−0.0800
	(−3.4181)	(1.8266)	(−2.1490)	(−0.7435)
OC	0.1937	0.5322 **	0.3942 *	−0.1382
	(0.8152)	(2.3255)	(1.7169)	(−0.8043)
FC	−2.6878 **	−1.6715 **	−2.6692 **	0.9034 **
	(−2.3734)	(−2.1627)	(−2.2094)	(2.4036)
Constant	18.2220 ***	5.5202 ***	7.2573 ***	5.9185 ***
	(15.9836)	(6.5929)	(5.7766)	(13.2614)
Firm FE	Yes	Yes	Yes	Yes
Year FE	Yes	Yes	Yes	Yes
N	4571	4559	4027	4571

Note: Columns (1) and (2) present the results when firm size was measured by total assets and number of employees, respectively, as the dependent variable. Column (3) shows the results when TFP was used as the dependent variable. Column (4) presents the results when supply chain efficiency was used as the dependent variable. The reduction in sample size was due to the need to replace product-level data with firm-level data when using firm-level variables as the dependent variable. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. The values in parentheses are t-values.

showed that the supply chain concentration had a significantly negative impact on firm size, total factor productivity, and supply chain efficiency. The impact on scale was likely because an increased supply chain concentration makes firms more dependent on major trading partners, thereby constraining the overall operational capacity and consuming resources, negatively affecting the firm size. When the firm size decreases, firms may find it harder to bear the high costs and long production cycles of high-quality export products, thus negatively impacting export product quality. Moreover, the impact on efficiency arises because as the supply chain concentration increases, firms’ activities become less flexible, and any issues with major trading partners can have a greater impact on firm operations, adversely affecting efficiency. When efficiency decreases, traditional economic theory suggests that reduced efficiency lowers the output per unit, possibly decreasing the elements that constitute the product, thus negatively impacting export product quality. Therefore, the supply chain concentration may reduce export product quality by shrinking the firm size and lowering efficiency, confirming Hypotheses 2 and 3.

5.3. Supply Chain Concentration, Resilience, and Export Product Quality

5.3.1. Infrastructure Resilience: Regional Fixed Asset Investment

When a region increases its fixed asset investment, areas such as transportation and research institutions may receive more funding, leading to further development. Furthermore, improved transportation conditions can enhance the firm’s operational efficiency, which relates to the channel through which the supply chain concentration negatively impacts export product quality. Therefore, better transportation infrastructure may affect the relationship between the supply chain concentration and export product quality. Moreover, when research institutions grow, they may provide more technological support to firms, which could also mitigate the negative effect of the supply chain concentration on firm efficiency, thereby influencing the relationship between the supply chain concentration and export product quality. In summary, it is worth investigating whether regional fixed asset investment influences the relationship between the supply chain concentration and export product quality. We obtained data on regional fixed asset investment ($\mathrm{RFAI}$) and analyzed it in natural logarithmic form. We constructed an interaction term between the supply chain concentration and regional fixed asset investment ($\mathrm{SCC}\times \mathrm{RFAI}$), then included both the univariate variable of regional fixed asset investment ($\mathrm{RFAI}$) and the interaction term ($\mathrm{SCC}\times \mathrm{RFAI}$) in the baseline model for regression analysis.

The results in column (1) of

Table 7. Supply chain concentration, resilience, and export product quality.

Variables	Dependent Variable: Quality
	(1)	(2)	(3)
SCC	−0.3214 ***	−0.0749 ***	−0.0776 ***
	(−5.0443)	(−4.5671)	(−5.3149)
RFAI	−0.0062
	(−1.4619)
SCC × RFAI	0.0300 ***
	(4.4195)
OS		−0.0261
		(−1.6330)
SCC × OS		0.1768 ***
		(3.7095)
IU			−0.0057
			(−1.1085)
SCC × IU			0.0284 **
			(2.4968)
Age	−0.0311 ***	−0.0125	−0.0317 ***
	(−4.6826)	(−1.2419)	(−4.7746)
SO	0.0025	−0.0099	0.0031
	(0.6695)	(−1.0798)	(0.8102)
Capital	0.0135 *	0.0223 *	0.0149 **
	(1.8326)	(1.9273)	(2.0179)
OC	−0.0434 ***	−0.0661 ***	−0.0446 ***
	(−3.5429)	(−4.1463)	(−3.6352)
FC	0.0410 ***	0.0453 **	0.0442 ***
	(2.6221)	(2.1903)	(2.8151)
Constant	0.7307 ***	0.6396 ***	0.6846 ***
	(15.3118)	(17.5874)	(26.5645)
Firm FE	Yes	Yes	Yes
Product FE	Yes	Yes	Yes
Destination FE	Yes	Yes	Yes
Year FE	Yes	Yes	Yes
N	397,002	248,974	396,663

Note: Columns (1) to (3) present the corresponding results of incorporating the regional fixed asset investment variable and its interaction term, the overseas subsidiaries variable and its interaction term, and the industrial upgrading variable and its interaction term into the baseline model, respectively. The lower number of observations in the second column is due to missing data on overseas subsidiaries. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively. The values in parentheses are t-values.

showed that the interaction term was significantly positive, indicating that higher regional fixed asset investment reduced the negative impact of the supply chain concentration on export product quality. This may be attributed to higher regional fixed asset investment improving infrastructure, such as transportation and research institutions, which can potentially enhance resilience, thereby helping firms mitigate the negative impacts of the supply chain concentration. When primary trading partners encounter issues, firms can leverage favorable transportation conditions to transport raw materials or products more efficiently, thereby better mitigating the negative impacts caused by issues in their primary trading relationships. Moreover, when supply chains fail to operate smoothly, well-developed research institutions can provide robust technical support, enabling firms to maintain high-quality export production amidst challenges, thereby reducing detrimental effects on export product quality.

5.3.2. Firm Structure Resilience: Overseas Subsidiaries

The establishment of overseas subsidiaries by firms may influence the supply chain concentration due to potentially improved procurement of overseas raw materials or exportation of products, hinting at potential research value. The results in column (2) of Table 7 showed that the interaction term was significantly positive, indicating that overseas subsidiaries may positively moderate the relationship between the supply chain concentration and export product quality. The overseas subsidiary variable ($\mathrm{OS}$) was measured by the proportion of overseas subsidiaries to total subsidiaries, following Lee et al. [55]. The positive moderating effect likely arose because, in the face of issues, a firm’s overseas subsidiaries may better support the firm by broadening its channels for sourcing raw materials or maintaining export activities, thus enhancing resilience. On one hand, when the supply chain concentration is high and major suppliers encounter problems, the firm can utilize its overseas subsidiaries to expand sourcing channels, mitigating the impact of supplier issues. On the other hand, when major clients encounter issues, overseas subsidiaries can assist in resolving these issues or seeking new partnerships with other clients, thereby reducing the impact of product stagnation on the financial stability of firms.

5.3.3. Industrial Structure Resilience: Industrial Upgrading

The regression results regarding the moderating effect of industrial structural upgrading are presented in Table 7, column (3). Industrial structural upgrading ($\mathrm{IU}$) was measured by the ratio of GDP from the tertiary sector to GDP from the secondary sector, drawing on the insights of Ceccobelli et al. [56]. The results indicated that the interaction term of industrial structural upgrading was significantly positive, suggesting that as industrial structures advance, the negative impact of the supply chain concentration on export product quality diminishes. The reason is that when the industrial structure becomes more advanced, the industrial arrangement in a region may become more reasonable [56], and hence better prepared to withstand the impacts of supply chain issues and recover more effectively. Moreover, when issues such as pandemics occur, the third sector industries, such as information technology services, can continue operations, and information technology can enhance operational efficiency for firms. Therefore, developing better information technology services can stabilize and effectively coordinate many activities during challenges, thereby assisting firms with greater resilience. This capability not only helps maintain operations within firms but also supports better cooperation with trading partners and facilitates assistance from non-trading partners during challenging times [57]. Consequently, industrial structural upgrading may mitigate the adverse effects of the supply chain concentration on export product quality.

6. Conclusions and Policy Implications

6.1. Conclusions

This study utilized data from Chinese A-share-listed companies in Shanghai and Shenzhen, along with customs data from 2001 to 2015, to examine the impact of the supply chain concentration on export product quality.

We found that an increase in the supply chain concentration reduced the quality of export products, and this conclusion held after robustness tests. In the heterogeneity analysis, focusing on aspects related to export product quality, namely, whether firms belonged to the high-tech industry and had banking connections, it was found that the negative impact of an increased supply chain concentration on export product quality was likely to be smaller when firms belonged to the high-tech industry and had connections with banks. Moreover, focusing on aspects related to the supply chain concentration, namely, whether the CEO had an overseas background and whether the top executives had expertise in digital-related fields, it was found that the negative impact of the supply chain concentration on export product quality was likely to be smaller when the CEO had an overseas background and the top executives had expertise in digital-related fields. Mechanism analysis revealed that an increased supply chain concentration may negatively affect export product quality by reducing the firm size, decreasing total factor productivity, and lowering supply chain efficiency. Finally, the moderation effect analysis of resilience revealed that regional fixed asset investment, overseas subsidiaries, and industrial upgrading, which respectively enhance infrastructure resilience, firm structure resilience, and industrial structure resilience, may all contribute to mitigating the adverse effects of an increased supply chain concentration on export product quality.

6.2. Policy Implications

The emergence of the COVID-19 pandemic has highlighted the significance of supply chain issues, giving this study’s findings practical relevance. The conclusions of this study suggested several policy implications: (1) Firms engaged in export trade should strive to avoid the concentration of their trading partners to reduce the negative impact of the supply chain concentration. (2) The government should pay closer attention to supply chain diversification in firms that do not belong to the high-tech industry, have no ties with banks, lack a CEO with an overseas background, and have top executives without a digital-related educational background. A reduction in supply chain diversification may have a more significant negative impact on these firms. (3) Since the supply chain concentration can adversely affect export product quality by reducing the firm size, total factor productivity, and supply chain efficiency, it is advisable to address these issues. Specifically, for firms with a high supply chain concentration, measures such as promoting collaborative operations among firms, conducting on-site learning activities with high-tech firms, and advancing the development of unmanned delivery technologies could be considered. (4) In regions where firms face relatively concentrated supply chains, the government should increase regional fixed asset investments, especially in transportation and research institutions, within the framework of well-planned government expenditure, to better help firms cope with supply chain risks. (5) When firms experience high levels of supply chain concentration, they should establish overseas subsidiaries to enhance their capability to expand overseas procurement channels and maintain product export operations when supply chain issues arise. This strategy aims to prevent the firm from bearing excessive losses when supply chain issues occur. (6) In planning industrial structural development, governments should, wherever feasible, prioritize expanding the tertiary sector to enhance the advancement of the industrial structure. Specifically, efforts should focus on promoting the development of information technology services and consider further advancements in AI technologies tailored for crisis response. Based on the above, these measures may be crucial for supporting firms in effectively addressing supply chain insecurity issues, thereby promoting sustainability.

6.3. Limitations and Future Work

This study has certain limitations. Firstly, the theoretical analysis in this paper could be more systematic if a theoretical model were constructed to examine the impact of changes in the supply chain concentration on firms and the resulting changes in export product quality. This represents an important limitation of the study. Secondly, due to data availability, export product quality based on customs data could only be accurately reflected up to 2015, which may not fully capture the most recent trends in export product quality. Thirdly, this research relied solely on data from listed companies, which may not fully represent the broader range of firms. In these two aspects, as data disclosure systems improve and datasets become more comprehensive in the future, subsequent research can leverage more complete datasets to better reflect the actual conditions of firms. Lastly, this paper has certain limitations in measuring industrial upgrading. The previous analysis only used a single indicator to assess industrial upgrading, and while some recent studies have adopted similar methods, literature that specifically focuses on industrial upgrading tends to use multiple indicators for a more systematic evaluation. As a result, the previous analysis may have limitations in comprehensively measuring industrial upgrading. Future research will consider drawing on existing literature for a more comprehensive and systematic assessment and will attempt to further refine the indicators based on evolving real-world conditions.

The previous analysis regarding research content may provide some reference for future studies. In this paper, we conducted a heterogeneity analysis by combining factors that are closely related to both the supply chain concentration and export product quality, revealing that these factors influence the relationship between the two. Therefore, future research can not only explore other factors with similar close ties but also conduct more detailed studies on each factor individually. Regarding the mechanism analysis, future studies could go beyond scale and efficiency to explore additional channels through which the supply chain concentration exerts its effects. This would contribute to a more comprehensive understanding of the pathways through which the supply chain concentration influences outcomes. Finally, the findings of this study suggested that enhancing resilience may help mitigate the negative impact of the supply chain concentration on export product quality. More importantly, resilience itself is critical to the functioning of supply chains, making further in-depth research into resilience highly valuable. Consequently, future research could not only focus on how to enhance the resilience of the supply chain itself but also explore the potential benefits of improving external resilience. This dual focus on both internal and external resilience would help better promote the healthy functioning of supply chains.