Supply Chain Stability and Enterprises’ Total Factor Productivity: From the Perspective of Development Sustainability

Abstract

Because of increasing global complexities and the frequent occurrence of black swan events, the risk of supply chain disruptions intensifies, and the sustainability of enterprise development faces significant challenges. By focusing on supply chain security and aiming for sustainable enterprise development, this study explores how stable supply chains contribute to enhanced productivity of core enterprises, using the theory of social capital as a framework. An empirical analysis was conducted using a sample of A-share listed companies in Shanghai and Shenzhen, China, from 2009 to 2021. The research findings indicate that greater supply chain stability is associated with higher total factor productivity in enterprises. The mechanism test results show that the supply chain stability relationship can improve the total factor productivity of enterprises through two paths, such as improving the operation and management level of enterprises and optimizing resource allocation. After further heterogeneity analysis, it is found that the stable relationship of the supply chain has a stronger positive impact on the total factor productivity of core enterprises in the non-state-owned enterprise group, the eastern enterprise group, and the large-size enterprise group. These findings help to provide insights for enterprises to achieve sustainable development in the new era.

1. Introduction

The reports from the 19th and 20th National Congresses of the Communist Party of China highlight a significant transition in China’s economy—from a phase of rapid growth to one focused on high-quality development. This shift in macroeconomic focus necessitates that micro enterprises also undergo high-quality development to provide the necessary support. However, in the current era, the sustainability of enterprise development faces significant challenges due to the frequent occurrence of black swan events. These include geopolitical conflicts, great power games, natural disasters, global public health crises, and so on. Such events create a complex and volatile external environment, complicating the efforts to ensure supply chain security and the sustainability of enterprise operations. The issue of national security has transcended the military sphere, and countries have taken measures to maintain supply chain security (On 1 January 2021, the EU Conflict Minerals Regulation came into force, making it mandatory for EU businesses not to procure or use raw materials of four metals—tungsten, tin, tantalum, and gold—and their manufactured products from conflict-affected or high-risk areas, and clarifying supplier due diligence obligations as an EU importer; on 9 August 2022, Biden signed the Chip and Science Act of 2022, which pushes for the manufacturing of its chips to be back to U.S. soil; on 16 March 2023, the European Commission adopted the EU Critical Raw Materials Act, which calls for diversification of EU supply by 2030 and reduces the EU’s dependence on imports from single-country suppliers; on 21 September 2023, the Chip in Europe Act came into force, driving investment in the chip sector in order to strengthen Europe’s technological inroads into semiconductors amidst a complex geopolitical backdrop leadership; and so on). In the context of “pan-national security”, it is worthwhile to think about how to balance development and security and take a sustainable path.

The twentieth report emphasized that “We should focus on advancing high-quality development and work diligently to increase total factor productivity”. As a key indicator of economic growth performance in enterprises, improving total factor productivity is crucial for fostering high-quality development and transforming the country’s economic growth model from “extensive” to “intensive” [1]. Enhancing the total factor productivity of enterprises has become a significant challenge and a prominent topic of interest today. And scholars have started their research on the external policy environment of enterprises [1], their own decision-making characteristics [2], and the characteristics of the supply chain in which the enterprises are located. At the supply chain level, López and Südekum (2009) demonstrated that, when considering vertical industry relationships, upstream industries exhibit significant cross-industry productivity spillover effects [3]. This finding suggests that upstream–downstream relationships within the supply chain can enhance overall sustainable development through productivity spillovers. Natividad (2014), through a case study of a large fishery company’s integrated management of its fish suppliers, found that knowledge transfer and hierarchical control within the supply chain significantly affect firms’ total factor productivity [4]. This indicates that the integration and management of supply chain relationships can facilitate knowledge sharing and resource allocation within the chain, which plays a crucial role in firms’ development. Fatima’s (2016) research revealed that foreign direct investment (FDI) through vertical supply channels positively influences the productivity levels of local firms [5]. This suggests that policymakers should strengthen the demand–supply relationships between local and multinational firms to optimize firm benefits, achieving mutual sustainability across both the supply chain and firms. Sari et al. (2016) further emphasized the importance of knowledge management within supply chain relationships for the sustainable development of firms within the chain [6]. Bagir and Seven (2022) found that an increase in financial constraints faced by upstream firms heightens the sensitivity of TFP growth to debt growth [7], implying that financial health within the supply chain is critical for productivity improvement and sustainability. Chen et al. (2024) showed that customers’ green technological innovations contribute to suppliers’ TFP [8], suggesting that technological innovation can propagate through supply chain relationships, thereby fostering productivity growth among chain-linked firms.

Together, these studies illustrate that supply chain linkages can exert a substantial impact on firms’ TFP through various pathways and that sustainable development within the supply chain is tightly intertwined with the sustainability of firm development. However, (1) most studies are one-way, based on either suppliers or customers. However, modern supply chains are essentially networked [9]. As a link in the supply chain network, enterprises are influenced by multiple customers and suppliers, making it necessary to examine the supply chain network from a global and holistic perspective. (2) There are fewer studies from the perspective of supply chain security. Previous studies have shown that disruptions in the supply chain can significantly negatively affect a firm’s financial, market, and operational performance [10]. He et al. (2022) argue that the formation of stable supply chain relationships around core firms helps to withstand competition and uncertainty disruptions, and promotes the improvement of supply chain performance and the realization of sustainable development [11]. It is evident that achieving stability in supply chain relationships supports the continuity of production and ensures the security needed to enhance the total factor productivity of enterprises. In addition, in a stable supply chain environment, enterprises are more likely to build mutually beneficial partnerships with their suppliers and customers, which is conducive to the collaborative development of chain members. So, can stable upstream and downstream relationships enhance the total factor productivity of core firms? This question requires further investigation.

This paper, set in the context of enhancing enterprise total factor productivity, conducts empirical research from the perspective of supply chain security to address the following questions: How does the stability of supply chain relationships impact enterprise total factor productivity? What are the pathways of this impact? Does this effect differ based on firm characteristics? Compared to the existing literature, the key contributions of this study are as follows:

(1)

Expanding the research on the factors influencing enterprise total factor productivity from the perspective of supply chain security, highlighting the crucial role of supply chain security in boosting the total factor productivity of core enterprises.

(2)

Offering a comprehensive framework for enterprises to maintain supply chain security and stability by focusing on both supply and demand aspects, emphasizing the importance of upstream and downstream partnerships within the supply chain network for maintaining dynamic stability.

(3)

Enriching the research on the economic outcomes of supply chain stability by revealing how it enhances enterprise total factor productivity through improved operational management and optimized resource allocation.

(4)

Providing clear guidance for enterprises of different ownership types, regions, and sizes. Non-state-owned enterprises should focus on building stable, long-term supply chain relationships to reduce uncertainty and ensure resource access and operational stability. For enterprises in China’s eastern region, leveraging the favorable economic environment and mature supply chains can enhance the positive impact of supply chain stability on productivity. Large-size enterprises should form strategic partnerships with supply chain partners to foster collaboration, enhance risk capacity, improve efficiency, and maximize scale advantages.

(5)

Drawing on social capital theory, this study explores the effect of supply chain stability on firms’ total factor productivity, thereby extending the application of social capital theory within the context of supply chain relationships and corporate management research.

The remainder of this study is structured as follows: Section 2 analyzes the theoretical pathways through which supply chain stability impacts firms’ total factor productivity and proposes corresponding research hypotheses. Section 3 describes the research design, including sample selection and data sources, model construction, and variable definitions. Section 4 presents the empirical results and analysis of the benchmark regression, endogeneity issues, and robustness tests. Section 5 conducts tests on the impact mechanisms and heterogeneity analysis. Section 6 summarizes the main findings and contributions of this study, along with practical implications.

2. Theoretical Analysis and Research Hypotheses

2.1. Supply Chain Stability and Total Factor Productivity of Enterprises

Firms’ total factor productivity results from the contribution of unproductive inputs, such as technological advancement and institutional improvements, to output growth [12]. The existing literature primarily examines the enhancement of enterprise total factor productivity through two key aspects: the effects of technological progress and resource allocation [13,14]. Technological progress includes two dimensions: production technology advancements and management technology improvements. Enterprises can increase total factor productivity by innovating production technologies and enhancing operational and management practices. The resource allocation effect involves optimizing the distribution of factors to reduce inefficient investments and costs while increasing efficient investments, thereby boosting total factor productivity. In the context of optimizing enterprise resource allocation, increasing R&D investment is a key criterion. It provides essential support for enterprises to engage in technological innovation activities, playing a vital role in fostering sustainable development [1].

As a comprehensive indicator reflecting the economic growth performance of enterprises, total factor productivity (TFP) is closely related to the operational status of its supply chain [15]. According to social capital theory, social capital resources are embedded in interconnected social networks, influencing the production operations and strategic decisions of enterprises [16]. The connections within social networks can be divided into strong ties and weak ties, which differ in interaction frequency, emotional intensity, intimacy, and reciprocal exchange [17]. Strong ties, characterized by trust, cooperation, and stability, can transmit high-quality, complex, or tacit knowledge and information. The close business interactions between enterprises and their suppliers and customers constitute a “strong tie” relationship, which is a valuable and inimitable resource. Enterprises in the supply chain obtain and optimize technical or capital resources through cooperation [18]. If the cooperative relationship between core enterprises and their upstream and downstream partners in the supply chain remains stable, it will form deep strong ties, enhance the motivation and willingness of suppliers and customers in the supply chain to deliver resources, and stabilize the core enterprises’ resource acquisition channels [19]. This helps enterprises obtain higher quality resources and information [20], thereby reducing information asymmetry, improving the level of integrated management and communication efficiency between upstream and downstream enterprises, reducing resource mismatches and cost losses, and enhancing the total factor productivity of enterprises. Accordingly, this paper proposes Hypothesis 1:

Hypothesis 1.

All else being equal, greater stability in the supply chain leads to higher total factor productivity in enterprises.

2.2. Technological Progress Effects

The effects of technological progress include not only advancements in production technology but also improvements in management levels, enhancement of competitive advantages through product innovation, and the application and promotion of green technologies, among other forms of progress. The stability of supply chain relationships primarily contributes to cooperation and coordination between partners in the upstream and downstream segments of the supply chain but does not necessarily directly drive technological research and innovation. The progress of an enterprise’s production technology, innovation, and similar areas is actually influenced by a mix of internal and external factors, such as capital and talent, organizational culture and institutional environment, development strategy, market demand and competitive pressure, policy support and regulatory environment, and external cooperation [21]. Therefore, there is no direct and obvious correlation between stable supply chain relationships and advancements in areas such as production technology. So, there is insufficient theoretical and literature support to consider factors such as production technology progress, etc., as intermediary factors between supply chain stability and enterprise total factor productivity.

The stability of supply chain relationships primarily influences enterprise total factor productivity (TFP) through improvements in operational management levels. On one hand, high information costs and information asymmetry are major causes of underinvestment and inefficiency in enterprises [22]. Stable partnerships between an enterprise and its upstream and downstream supply chain partners effectively establish a robust information-sharing channel [19]. According to social capital theory, relational social capital focuses on the specific relationships and interactions between individuals or groups, which form the foundation of trust and reciprocity. Trust among supply chain members enhances their willingness to share and accept information, thereby facilitating the better transmission of both “hard information” and “soft information” within internal networks [23]. This effectively improves communication quality and efficiency, aligns information between the core enterprise and its suppliers and customers, reduces information distortion, alleviates the “bullwhip effect” in the supply chain, and enhances TFP [24]. On the other hand, the communication platforms and trust levels built on stable supply chain relationships [23], as well as increased coordination among supply chain members [25], make it more efficient and effective for the core enterprise to manage its supply chain. This drives the advancement of integrated management technologies for upstream and downstream operations, providing technical assurance for improved supply chain management efficiency and further promoting TFP. Accordingly, this paper puts forward Hypothesis 2a:

Hypothesis 2a.

All else being equal, stable supply chain relationships can enhance firms’ total factor productivity by improving the level of operational management.

2.3. Resource Allocation Effects

Through trust and reciprocal relationships established with other firms in the supply chain network (e.g., customers, suppliers), companies can access higher-quality structural social capital, cognitive social capital that facilitates coordinated actions and collaboration, and relational social capital that supports resource acquisition and sharing. These forms of social capital aid in optimizing resource allocation, thereby further enhancing firm performance. On the one hand, enhanced supply chain stability aids in minimizing associated costs and expenses. Enterprises typically need to allocate resources for gathering information about suppliers and customers and maintaining these relationships. Bourdieu (1986) posits that the essence of an enterprise’s ability to obtain resources through social networks lies in the costs incurred in building these relationship networks—the poorer the supply chain stability, the higher the costs [26]. If upstream and downstream relationships remain stable, the efficiency of information transmission and the mitigation of the “bullwhip effect” will significantly reduce resource waste caused by information asymmetry in the supply chain [27]. Furthermore, deeper trust and cooperation between partner enterprises will foster integrated “strategic relationships,” reducing opportunistic behavior among suppliers and customers, and lowering transaction uncertainty and negotiation frequency, thus decreasing transaction costs and supply chain switching costs [19]. Additionally, stable supply chain relationships can strengthen an enterprise’s bargaining power, thereby minimizing losses during negotiations [28].

On the other hand, greater supply chain stability can increase an enterprise’s investment in innovation. Hilary and Hui (2009) argue that fluctuations in supply chain relationships can diminish incentives for enterprise innovation, whereas stable supply chain relationships are essential for securing bank and commercial credit financing [29]. Enhanced supply chain stability reduces financing constraints [20], thus promoting rational fund allocation, increasing investment in innovation, and optimizing resource allocation. Accordingly, this paper proposes Hypothesis 2b:

Hypothesis 2b.

All else being equal, stable supply chain relationships can enhance firms’ total factor productivity by optimizing resource allocation.

To empirically validate the aforementioned hypotheses, the research design flowchart is presented in Figure 1.

3. Empirical Research Design

3.1. Sample Selection and Data Sources

Since more firms have voluntarily disclosed supplier and customer information only since 2009, this paper uses A-share listed firms in Shanghai and Shenzhen, China, as the initial sample for the period from 2009 to 2021. After excluding samples in the financial industry, ST, *ST, and PT, those with missing data on key variables, and those with abnormal or missing financial data, the final research sample contains 13,849 firm–year observations. Data on supply chain relationships, financial metrics, and firm characteristics for the listed companies were obtained from the CSMAR and CNRDS databases. To mitigate the impact of outliers, all continuous variables in this study were winsorized at the 1% level at both ends. Econometric analyses were conducted using Stata16 software.

3.2. Modeling

To examine the impact of supply chain stability on firms’ total factor productivity, drawing on Wang and Han’s (2023)’s study [30], this paper sets up the benchmark econometric model as follows:

$${TFP}_{it}={\mathsf{\alpha }}_0+{\mathsf{\alpha }}_1{Chainst}_{it}+{\mathsf{\alpha }}_2{Controls}_{it}+{\mu }_i+{\theta }_t+{\epsilon }_{it}$$

(1)

where the explained variable ${TFP}_{it}$ denotes the total factor productivity of core firms, the explanatory variable ${\mathrm{C}\mathrm{h}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{s}\mathrm{t}}_{\mathrm{i}\mathrm{t}}$ denotes the stability of upstream and downstream supply chains of core firms, and the set of control variables is represented by ${Controls}_{it}$. ${\alpha }_0$, ${\alpha }_1$, and ${\alpha }_2$ are the regression coefficients to be estimated. The firm fixed effects are denoted by ${\mu }_i$, the year fixed effects are denoted by ${\theta }_t$, and ${\epsilon }_{it}$ represents the random error term. Among them, the subscripts i and t indicate the corresponding company and year.

3.3. Definition of Variables

3.3.1. Explained Variable: Total Factor Productivity of Enterprises

The existing literature identifies five main methods for measuring the total factor productivity (TFP) of firms: generalized least squares (GLS), fixed effects (FEs), Olley–Pakes (OP), Levinsohn–Petrin (LP), and generalized method of moments (GMM). Among these, the OP, LP, and GMM methods are particularly effective in addressing endogeneity issues caused by variable interdependence and sample selection bias. The LP method, in particular, is advantageous in reducing sample loss [12]. Therefore, this paper employs the LP method to measure firms’ total factor productivity (Tfp_lp) in the benchmark regression and uses the OP and GMM methods (Tfp_op, Tfp_gmm) for robustness testing.

$${\mathrm{Y}}_{\mathrm{i}\mathrm{t}}={\mathrm{A}}_{\mathrm{i}\mathrm{t}}{\mathrm{L}}_{\mathrm{i}\mathrm{t}}^{\mathsf{\alpha }}{\mathrm{K}}_{\mathrm{i}\mathrm{t}}^{\mathsf{\beta }}$$

(2)

$$\mathrm{l}\mathrm{n}{\mathrm{Y}}_{\mathrm{i}\mathrm{t}}=\mathsf{\alpha }\mathrm{l}\mathrm{n}{\mathrm{L}}_{\mathrm{i}\mathrm{t}}+\mathsf{\beta }\mathrm{l}\mathrm{n}{\mathrm{K}}_{\mathrm{i}\mathrm{t}}+{\mathsf{\mu }}_{\mathrm{i}\mathrm{t}}$$

(3)

Specifically, we use a Cobb–Douglas production function to specify the form of the production function (Equation (2)), where Y represents output, A denotes total factor productivity (TFP), and L and K indicate labor and capital inputs, respectively (Y = operating revenue/10,000; L = number of employees; K = net fixed assets/10,000). By taking the logarithm of both sides of Equation (2), we can transform it into the linear form shown in Equation (3). The residual term in Equation (3) contains information about the logarithmic form of firms’ TFP, which we estimate.

To address simultaneity bias and sample selection bias in the estimation process, the OP method assumes that firms make investment decisions based on current productivity. Accordingly, we use a firm’s current investment (I) (I = cash paid for the purchase and construction of fixed assets, intangible assets, and other long-term assets /10,000) as a proxy variable for unobserved productivity to correct for simultaneity bias. For sample selection bias, the OP method constructs a polynomial including the firm’s investment and capital stock logarithms to obtain a consistent and unbiased estimate of labor output elasticity. It then builds a firm survival probability model to estimate firms’ entry and exit decisions, effectively resolving sample selection bias. The LP method, building on the OP method’s assumptions, replaces the proxy variable of current investment with intermediate inputs (M) (M = (operating costs + selling expenses + administrative expenses + financial expenses − depreciation and amortization − cash paid to and on behalf of employees)/10,000). The GMM method, in turn, adds the lagged log of ${\mathrm{Y}}_{\mathrm{i}\mathrm{t}}$ as an instrumental variable to address endogeneity in the model, enabling estimation of the production function.

3.3.2. Explanatory Variable: Supply Chain Stability

Drawing on Bernard et al.’s (2010) measurement approach for product switching in trade research [31], Jiang and Lu (2022) [32] use the number of new suppliers in the top five supplier list at the end of the year, compared to the previous year, divided by 5 to measure supplier turnover. Similarly, for clients, they use the number of new clients in the top five client list at the end of the year, compared to the previous year, divided by 5 to measure client turnover. The average of the supplier and client turnover rates (i.e., the total number of changes in the top five upstream and downstream trading partners divided by 10) is taken as an indicator of supply chain relationship turnover.

Following Jiang and Lu’s measurement of supply chain relationship turnover [32], this paper first measures supplier stability by dividing the number of suppliers in the year-end top five supplier list that overlap with the previous year by 5, and client stability by dividing the number of clients in the year-end top five client list that overlap with the previous year by 5. Then, the product and average of supplier stability and client stability (the average being the total number of overlaps in the top five upstream and downstream trading partners compared to the previous year, divided by 10) are used as two indicators (Chainst1 and Chainst2) to measure supply chain stability in this empirical analysis.

3.3.3. Control Variable

To enhance the accuracy of the study, control variables from the existing literature, as listed in

Table 1. Definition of main variables and descriptive statistics.

Variable Type	Variable Name	Variable Symbol	Variable Definition	N	Mean	SD
Implicit Variable	Total factor productivity of enterprises	Tfp_lp	total factor productivity calculated by the lp method	13,849	8.360	0.990
Independent Variable	Supply chain stability	Chainst1	supplier stability × customer stability	13,849	0.660	0.440
		Chainst2	percentage of supplier and customer duplication compared to the prior year	13,849	0.730	0.370
Control Variable	Enterprise size	Size	the total assets of the enterprise are taken as the natural logarithm	13,849	22.18	1.160
	Age of business	Age	the number of years the company has been listed plus one takes the natural logarithm	13,849	2.230	0.620
	Return on assets	Roa	net profit/total assets closing balance of the previous year × 100	13,849	4.310	7.800
	Gearing	Lev	total liabilities/total assets	13,849	0.420	0.200
	Percentage of independent directors	Indrcrat	proportion of independent directors to the total number of all directors	13,849	0.380	0.100
	Board size	Board	the number of directors on the board of directors is taken as the natural logarithm	13,849	2.240	0.290
	Shareholding ratio of top five shareholders	Top5	sum of shareholdings of the company’s top five largest shareholders	13,849	0.510	0.150
	Growth	Growth	(current period − prior year’s adjustments)/ABS prior year’s adjustments	13,849	15.65	34.29
	Fixed asset ratio	Tang	fixed assets/total assets	13,849	0.210	0.150
	Current asset ratio	Qr	current assets/current liabilities	13,849	2.320	2.090
	Whether or not the two posts are combined	Dual	the positions of chairman and general manager are assigned a value of 1 when combined; otherwise, it is 0	13,849	0.310	0.460

, were included in this paper. Table 1 provides the definitions of the main variables and their descriptive statistics.

4. Empirical Results and Analyses

4.1. Benchmark Regression

Table 2. Benchmark regression of supply chain stability and firms’ total factor productivity.

	(1)	(2)	(3)	(4)
	Tfp_lp	Tfp_lp	Tfp_lp	Tfp_lp
Chainst1	0.027 ***		0.036 ***
	(3.536)		(4.896)
Chainst2		0.027 ***		0.040 ***
		(3.324)		(4.966)
Size	0.483 ***	0.483 ***	0.537 ***	0.537 ***
	(61.932)	(61.982)	(85.666)	(85.741)
Age	0.108 ***	0.108 ***	0.091 ***	0.091 ***
	(4.857)	(4.848)	(7.259)	(7.219)
Lev	0.236 ***	0.237 ***	0.308 ***	0.309 ***
	(7.000)	(7.010)	(9.796)	(9.820)
Indrcrat	−0.012	−0.011	−0.018	−0.018
	(−0.371)	(−0.356)	(−0.588)	(−0.570)
Growth	0.002 ***	0.002 ***	0.002 ***	0.002 ***
	(29.162)	(29.167)	(27.387)	(27.410)
Board	−0.011	−0.011	−0.008	−0.008
	(-0.925)	(-0.939)	(−0.664)	(−0.678)
Tang	−1.166 ***	−1.167 ***	−1.062 ***	−1.063 ***
	(-28.214)	(-28.240)	(-28.428)	(-28.457)
Qr	−0.013 ***	−0.013 ***	−0.016 ***	−0.016 ***
	(−5.314)	(−5.295)	(−6.907)	(−6.886)
Roa	0.008 ***	0.008 ***	0.009 ***	0.009 ***
	(16.571)	(16.570)	(19.457)	(19.471)
Dual	−0.005	−0.005	−0.011	−0.011
	(−0.612)	(−0.593)	(−1.373)	(−1.344)
Top5	0.036	0.037	0.127 ***	0.127 ***
	(0.776)	(0.791)	(3.184)	(3.198)
_cons	−2.415 ***	−2.426 ***	−4.002 ***	−4.014 ***
	(−14.409)	(−14.478)	(−26.164)	(−26.254)
ID	Yes	Yes
Year	Yes	Yes	Yes	Yes
Province			Yes	Yes
Industry			Yes	Yes
N	13,849	13,849	13,849	13,849
R2	0.590	0.590	0.602	0.602

Note: Significance at the 1% levels is indicated by ***, with t-values presented in parentheses beneath the regression coefficients.

presents the benchmark regression results based on model (1), examining the impact of supply chain stability on firms’ total factor productivity. In columns (1) and (2), the regression coefficients for the core explanatory variables, Chainst1 and Chainst2, which represent supply chain stability’s effect on total factor productivity (Tfp_lp), are both 0.027. These coefficients are statistically significant and positive at the 1% level, controlling for year and firm fixed effects. When the clustering hierarchy is changed from firm–year to industry–province–year, as shown in columns (3) and (4), the regression coefficients for Chainst1 and Chainst2 increase to 0.036 and 0.040, respectively, and remain statistically significant at the 1% level. This demonstrates that enhancing supply chain stability can significantly boost the total factor productivity of enterprises, providing empirical support for Hypothesis 1.

4.2. Endogeneity Issues and Robustness Tests

4.2.1. Endogenous Issues

(1)

Heckman Two-Stage Regression

Given that listed companies have the discretion to disclose supplier and customer information, this paper employs the Heckman two-step approach to address potential sample selection bias, drawing on Ellis et al. (2012) [33]. In the first stage, a probit regression is conducted where the disclosure of supplier or customer information (disclosure) is the dependent variable (taking a value of 1 if disclosed, and 0 otherwise). Explanatory variables include firm size, age, gearing ratio, growth, return on assets, and the nature of the firm. The Inverse Mills Ratio (IMR) estimated in the first stage is included in the second-stage regression model. Column (1) of

Table 3. Endogeneity test.

	(1)	(2)	(3)
	Tfp_lp	Chainst1	Tfp_lp
Chainst1	0.115 ***		0.387 **
	(9.945)		(1.994)
Mean_YP		0.099 *
		(1.835)
Mean_YA		0.496 ***
		(3.556)
IMR	0.074 ***
	(2.596)
CVs	Yes	Yes	Yes
ID	Yes	Yes	Yes
Year	Yes	Yes	Yes
N	13,849	13,849	13,849
F		11.720	15.880
Sargan-P			0.614
A LM-P			0.000

Note: Significance at the 10%, 5%, and 1% levels is indicated by *, **, and ***, respectively, with z-values in parentheses under the regression coefficients in column (1) and t-values in parentheses under the regression coefficients in columns (2) and (3).

presents the results of the second-stage regression. The significance level of the IMR coefficient is at 1%, indicating that sample selection bias is indeed present in the original equation. The regression coefficient for the explanatory variable Chainst1 is statistically significant and positive at the 1% level, suggesting that the conclusions of the benchmark regression remain valid even after accounting for sample selection bias.

(2)

Instrumental Variable Approach

Model (1) accounts for various firm-level characteristics and fixed effects, such as individual and time effects, but unobservable factors may still interfere. Additionally, causal identification using the regression model may face endogeneity issues due to reverse causation, meaning that firms with high total factor productivity are better at maintaining supply chain stability. To address endogeneity issues like omitted variables and reverse causation, this paper employs a two-stage instrumental variable approach, following the methodology of Xu et al. (2014) [34]. It selects the mean values of other firms in the same province and region in the same year (Mean_YP and Mean_YA) as instrumental variables for supply chain stability in the 2SLS estimation. The results are reported in columns (2) and (3) of Table 3.

In the first-stage regression, the sign of the instrumental variables is significantly positive, indicating that they are strong predictors of the explanatory variables. The F-statistic for the first-stage regression is 11.720, which is greater than 10, allowing us to reject the null hypothesis of weak instruments. The Sargan test for Chainst1’s instrumental variables yields a p-value of 0.614, greater than 0.1, indicating no over-identification problem and proper selection of the instruments. In the under-identification test, the Anderson LM statistic has a p-value of 0.000, less than 0.01, significantly rejecting the null hypothesis of “under-identification of instrumental variables” at the 1% level. Furthermore, the second-stage results show that the coefficient of Chainst1 is significantly positive at the 5% level, confirming that the previous conclusions hold even after addressing endogeneity.

4.2.2. Robustness Check

To verify the robustness of the previous findings, this paper implements the following tests based on the benchmark regression model (1):

(1)

Alternative Methods: We re-calculate firms’ total factor productivity (Tfp_op, Tfp_gmm) using two alternative methods: the Olley–Pakes (OP) method and the generalized method of moments (GMM), as shown in columns (1) and (2) of Table 4.

(2)

Adjusting for COVID-19 Impact: Due to the significant disruption of global supply chains and business operations caused by the COVID-19 pandemic in 2020, we shorten the sample period and exclude data from 2020 onwards, as indicated in column (3) of Table 4.

(3)

Outlier Treatment: We further winsorize all continuous variables at the 2% and 5% levels, respectively, to minimize the impact of outliers, as detailed in columns (4) and (5) of Table 4.

Table 4. Robustness test.

	(1)	(2)	(3)	(4)	(5)
	Tfp_op	Tfp_gmm	tfp_lp	Tfp_lp	Tfp_lp
Chainst1	0.027 ***	0.030 ***	0.032 ***	0.023 ***	0.018 **
	(3.581)	(3.557)	(3.614)	(3.147)	(2.529)
CVs	Yes	Yes	Yes	Yes	Yes
ID	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes
N	13,849	13,849	9791	13,849	13,849
R2	0.504	0.411	0.578	0.592	0.586

Note: Significance at the 5% and 1% levels is indicated by ** and ***, respectively, with t-values presented in parentheses beneath the regression coefficients.

Table 4. Robustness test.

	(1)	(2)	(3)	(4)	(5)
	Tfp_op	Tfp_gmm	tfp_lp	Tfp_lp	Tfp_lp
Chainst1	0.027 ***	0.030 ***	0.032 ***	0.023 ***	0.018 **
	(3.581)	(3.557)	(3.614)	(3.147)	(2.529)
CVs	Yes	Yes	Yes	Yes	Yes
ID	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes
N	13,849	13,849	9791	13,849	13,849
R2	0.504	0.411	0.578	0.592	0.586

Note: Significance at the 5% and 1% levels is indicated by ** and ***, respectively, with t-values presented in parentheses beneath the regression coefficients.

5. Further Analysis

5.1. Impact Mechanism Testing

The previous section confirms the positive impact of supply chain stability on firms’ total factor productivity. The following section further explores the mechanisms driving this effect.

5.1.1. Supply Chain Stability, Operations Management Level, and Enterprise Total Factor Productivity

According to the above, the intermediary role of operation management level is mainly reflected in two aspects, namely supply chain information transfer efficiency and management technology. Firstly, high-quality and efficient transfer of supply and market information can facilitate upstream and downstream integration within the supply chain. This integration enhances an enterprise’s control over various aspects of the product life cycle, allowing it to more accurately predict and plan for product demand, production, and disposal. As a result, it improves the precision of asset allocation and increases the fixed asset turnover rate. Therefore, in this paper, the upstream and downstream information transfer efficiency of the enterprise is measured by the enterprise’s fixed asset turnover rate (Fxastto), which is defined as total operating income/average net fixed assets. In addition, digital technologies such as artificial intelligence, big data, cloud computing, and so on can provide efficient upstream and downstream integration management technology means for the core enterprise, so this paper measures the enterprise management technology by the level of digitalization [2]. Drawing on Wang and Han et al. (2023) [30], searching, matching, and word frequency counting based on the feature words of big data, blockchain, AI, cloud computing, and other modules, and then categorizing and grouping the word frequencies of key technological directions and forming the final summed word frequencies, and then logarithmically processing them, allows us to obtain the overall index that portrays the digital level of the enterprise (Digitala).

Drawing on Wen et al.’s (2004) [35] proposed mediation effect testing procedure, the mediation mechanism is evaluated through the following steps: First, we assess the regression coefficients of the explanatory variable Chainst1 on the mediating variables Fxastto and Digitala. As shown in columns (1) and (3) of

Table 5. Intermediary mechanism test.

	Operations Management Level				Resource Allocation
	(1)	(2)	(3)	(4)	(5)	(6)
	Fxastto	Tfp_lp	Digitala	Tfp_lp	RD1	Tfp_lp
Fxastto		0.001 ***
		(13.187)
Digitala				0.001 ***
				(5.368)
RD1						6.852 ***
						(23.245)
Chainst1	3.351 **	0.025 ***	1.339 ***	0.026 ***	0.001 **	0.023 ***
	(2.407)	(3.311)	(2.722)	(3.411)	(2.272)	(3.076)
CVs	Yes	Yes	Yes	Yes	Yes	Yes
ID	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
N	13,849	13,849	13,849	13,849	13,849	13,849
R2	0.009	0.597	0.095	0.591	0.072	0.613

Note: Significance at the 5% and 1% levels is indicated by ** and ***, respectively, with t-values presented in parentheses beneath the regression coefficients.

, the coefficient of Chainst1 is statistically significantly positive at least at the 5% level, indicating that a more stable supply chain relationship enhances the efficiency of information transfer and digital management technology in firms, thus improving their operational and management capabilities. Second, the mediating variables Fxastto and Digitala are added to the regression along with the explanatory variable Chainst1. The results in columns (2) and (4) of Table 5 indicate that both Chainst1 and the mediating variables Fxastto and Digitala have significantly positive coefficients. This suggests that firms’ digital management technology and the efficiency of supply chain information transfer partially mediate the effect of supply chain stability on firms’ total factor productivity. In essence, stable supply chain relationships can improve an enterprise’s total factor productivity by enhancing its operational and management capabilities, confirming H2a.

5.1.2. Supply Chain Stability, Resource Allocation, and Total Factor Productivity of Firms

The optimization of resource allocation efficiency essentially refers to the reallocation of enterprise resources from less efficient to more efficient projects [1], which can be measured through reduced operating costs, decreased waste, and increased investment in high-return projects. Numerous empirical studies have demonstrated that stable supply chain relationships can lead to reduced costs and expenditures [19,25,27,28]. Meanwhile, as R&D and innovation capabilities are core competencies for enterprise development, increased investment in R&D and innovation has become a key indicator in resource allocation optimization. Therefore, based on the verified negative effect of supply chain stability on corporate costs and expenses, this paper further tests whether stable supply chain relationships positively impact investment in high-efficiency projects. This allows us to examine whether resource allocation optimization serves as a mediating factor between supply chain stability and enterprise total factor productivity. The increase in investment in high-efficiency projects is measured through an increase in R&D and innovation investments. Drawing on Zhuo and Chen (2023) [21], this study uses the proportion of annual R&D expenditure to total assets as a measure of enterprise innovation R&D investment. Then, the mediation mechanism is tested as follows.

First, we test the regression coefficients of the explanatory variable Chainst1 on the mediating variable RD1. The results in column (5) of Table 5 indicate that the coefficient of Chainst1 is significantly positive at the 5% level, suggesting that stable supply chain relationships promote increased innovation investment by firms. Second, the mediating variable RD1 is added to the regression alongside Chainst1. The results in column (6) of Table 5 show that both Chainst1 and RD1 have significantly positive coefficients, indicating that innovation inputs partially mediate the impact of supply chain stability on firms’ total factor productivity. In other words, stable supply chain relationships help firms reduce related costs and increase investment in high-efficiency projects, meaning they can enhance firms’ total factor productivity through optimized resource allocation, thereby confirming H2b.

5.2. Heterogeneity Analysis

Generally speaking, the nature of the enterprise itself, its size, and the degree of economic development of the region in which it is located are closely related to the enterprise’s risk-coping ability and the degree of reliance on a stable supply chain, so this paper will analyze the heterogeneity from the following three levels.

5.2.1. Heterogeneity Analysis Based on the Nature of Business Ownership

Compared to state-owned enterprises, non-state-owned enterprises lack government resources and implicit guarantees [36], which exposes them to a more uncertain business environment and weaker risk-taking capacity [37]. As a result, maintaining stable supply chain relationships is especially crucial for non-state-owned enterprises, as it ensures resource acquisition and operational stability. Consequently, this paper hypothesizes that the positive effect of supply chain stability on total factor productivity will be more pronounced in non-state-owned enterprises.

In this paper, the samples are divided based on the ownership nature of the firms (OW = 1 is state-owned firms and OW = 0 is non-state-owned firms). The results of these grouping tests are presented in columns (1) and (2) of

Table 6. Heterogeneity test to distinguish nature, region, and size of firms.

	(1)	(2)	(3)	(4)	(5)	(6)
	OW = 0	OW = 1	Eastern	Central–Western	Large-Size	Small-Size
Chainst1	0.033 ***	0.003	0.030 ***	0.023	0.029 **	0.015
	(3.684)	(0.184)	(3.354)	(1.601)	(2.532)	(1.550)
CVs	Yes	Yes	Yes	Yes	Yes	Yes
ID	Yes	Yes	Yes	Yes	Yes	Yes
Year	Yes	Yes	Yes	Yes	Yes	Yes
N	9949	3900	9978	3871	6139	7710
R2	0.603	0.569	0.583	0.606	0.503	0.494

Note: Significance at the 5% and 1% levels is indicated by ** and ***, respectively, with t-values presented in parentheses beneath the regression coefficients.

. The coefficient of the explanatory variable Chainst1 is positive for both groups; however, it is statistically significant at the 1% level for non-state-owned enterprises, while it is not significant for state-owned enterprises. This suggests that the positive impact of supply chain stability on total factor productivity is more pronounced in non-state-owned enterprises compared to state-owned enterprises.

5.2.2. Heterogeneity Analysis Based on the Region to Which the Firm Belongs

Generally speaking, compared with the eastern region, the overall degree of marketization in the central and western regions of China is lower, and the possibility of opportunistic behavior among market players is higher [38]. As a result, the vulnerability of supply chains in the central and western regions forces chain firms to pay attention to the establishment and maintenance of stable relationships in the supply chain. In contrast, in the eastern region, where the degree of marketization is higher, the market rules are more mature, the upstream and downstream enterprises have abundant resources, and the contribution of supply chain stability to the total factor productivity of the enterprises in the eastern region seems to be less than that of the enterprises in the central and western regions.

Drawing on existing studies [39], this paper categorizes sample provinces and regions into eastern and central–western groups based on geographical location and economic development levels (The central and western groups include Hubei, Henan, Anhui, Inner Mongolia, Sichuan, Hunan, Shanxi, Jiangxi, Guangxi, Jilin, Guizhou, Yunnan, Chongqing, Shaanxi, Heilongjiang, Gansu, and Qinghai and Xizang, Ningxia, and Xinjiang. The eastern group includes Tianjin, Jiangsu, Liaoning, Fujian, Shandong, Zhejiang, Shanghai, Beijing, Guangdong, Hebei, and Hainan). The results of the grouping test are presented in columns (3) and (4) of Table 6. Contrary to expectations, the coefficient of the explanatory variable Chainst1 is insignificantly positive in the central–western group but significantly positive in the eastern group. This indicates that supply chain stability has a greater positive impact on enterprise total factor productivity in the eastern region compared to the central–western regions. This difference may be attributed to the higher levels of economic development and technological and managerial efficiency in the eastern region, which amplify the effects of supply chain stability on total factor productivity. Additionally, policy support and preferences in the central–western regions may diminish the impact of supply chain stability on firms’ total factor productivity, despite a lower level of marketization.

5.2.3. Heterogeneity Analysis Based on Firm Size

Compared to small-size firms, large-size firms have more resources and specialized assets and enjoy higher voice and bargaining power upstream and downstream in the supply chain. The larger the size of the core firm and the higher the degree of inequality in the supply chain, the more dominant the core enterprise is [40]. In this case, upstream and downstream partners will do their best to ensure the supply of raw materials or product sales, and a stable supply chain relationship will have a stronger effect on the total factor productivity of the enterprise. In addition, enterprise size is significantly negatively correlated with risk-taking; small-scale firms are more risk-averse in their choice of investment projects, while large-scale firms need supply chain stability to ensure their smooth operation as they interface with larger supply and demand and face more complex operation management [41]. Therefore, this paper hypothesizes that supply chain stability has a more significant positive impact on total factor productivity in large-scale enterprises.

The samples are divided based on the median enterprise size: firms with a size value greater than the median are classified as large-size, while those with a value less than or equal to the median are classified as small-size. The results of the grouping test are presented in columns (5) and (6) of Table 6. These results demonstrate that the positive impact of supply chain stability on total factor productivity is significantly stronger in large-size enterprises compared to small-size enterprises, confirming the prediction of this paper.

6. Conclusions and Insights

This paper, grounded in the contemporary context and based on social capital theory, introduces the perspective of supply chain security into the study of firms’ total factor productivity, exploring sustainable development paths for current enterprises. Specifically, using data from 2009 to 2021, it analyzes the impact of upstream and downstream supply chain stability of listed companies on their total factor productivity. Through rigorous analysis involving baseline regression, endogeneity analysis, and robustness checks, this paper provides substantial support for Hypothesis 1, finding that stable supply chain relationships significantly enhance firms’ total factor productivity (TFP). In further analyses, the study investigates the heterogeneous effects of supply chain stability on TFP across three dimensions: ownership type, geographic region, and firm size. The results indicate that the positive impact of supply chain stability on core firms’ TFP is more pronounced among non-state-owned firms, firms located in eastern regions, and larger firms. Mediation mechanism tests further validate Hypotheses H2a and H2b, confirming that the positive effect of supply chain stability on TFP operates through two channels: improvements in operational management and optimization of resource allocation.

This study is conducted within the context of firms’ external risk environments and China’s economic development, presenting theoretical and practical innovations in comparison to the previous literature as follows: (1) It enriches research on high-quality economic development from a micro-level perspective. (2) It expands the study of factors influencing firms’ high-quality development from a supply chain security perspective, providing practical insights into drivers of high-quality development aligned with the current era. (3) It advances research on the economic outcomes of supply chain stability, revealing the underlying mechanisms and variations in its effects. (4) It offers a comprehensive approach for firms to maintain the security and stability of their supply chains. While prior studies often focus on single-sided supply chain relationships, this study, aligned with the evolution of supply chain networks, examines both the supply and demand sides, underscoring the importance of maintaining dynamic stability across upstream and downstream partnerships. (5) It broadens the focus to include the long-term, dynamic aspects of relationships between firms and supply chain stakeholders, highlighting the impact of these sustained relationships on firms’ sustainable development. (6) Finally, by examining the impact of supply chain stability on firms’ total factor productivity based on social capital theory, this study enhances the application of social capital theory in the contexts of supply chain relationships and corporate management research.

Additionally, this study offers the following practical insights for promoting high-quality development of firms and the economy in the current era:

(1)

Enterprises need to establish an awareness of supply chain security and strive to build safe and reliable supply chain networks, promoting development through security and pursuing sustainable development paths. Supply chain risks are transmissive, with safety issues affecting the entire chain. The pursuit of short-term economic benefits by enterprises today is unadvisable. For enterprises to thrive in risk environments and attain high-quality long-term development, they must build strong partnerships within their upstream and downstream networks and improve the efficiency and stability of transferring both explicit and implicit resources between firms. Additionally, stable supply chain relationships do not mean static ones; rather, core enterprises must maintain dynamic stability in their upstream and downstream relationships based on their strategic development needs. This requires firms to continuously enhance their core competitiveness and supply chain discourse power, thereby having the ability to build stable supply chain relationships and choose partners.

(2)

Based on the mechanism testing results of this study, firms should focus on effective communication and integrated management with upstream and downstream enterprises in their daily operations, continuously optimizing resource allocation, reducing redundant costs, and increasing innovation investment to enhance economic efficiency. Furthermore, firms should manage supply chain risks based on their specific characteristics. Non-state-owned enterprises, those located in eastern regions, and large-scale enterprises should leverage the stability of supply chain relationships to enhance their total factor productivity and drive sustainable development.

(3)

Governments must understand the urgency of industrial and supply chain security with a bottom-line thinking approach, making the balance between development and security the main theme of economic governance. In the international environment, great power competition, wars, geopolitical issues, and public health crises threaten global industrial and supply chain security. Governments, as administrative bodies, should actively deepen international cooperation on industrial and supply chain security and stability, safeguard the common interests of countries, and firmly oppose the trend of “de-globalization.” Domestically, governments should refine industrial policies related to secure development and enhance governance capabilities for industrial and supply chain security. By creating a favorable development environment for market entities and accelerating the establishment of a new development pattern that prioritizes “domestic circulation as the mainstay while mutually reinforcing domestic and international circulations,” governments can support enterprises in achieving sustainable development.