The Impact of Digital Transformation on Supply Chain Resilience in Manufacturing: The Mediating Role of Supply Chain Integration

Abstract

The national economy’s steady flow can be ensured by strengthening supply chain resilience, and one of the main factors influencing supply chain resilience is business digital transformation. This study examines whether business digital transformation affects supply chain resilience using the panel data of Chinese listed manufacturing companies from 2012 to 2022. The findings indicate that supply chain resilience could be greatly enhanced by the digital transformation of manufacturing firms. The mediation mechanism test results show that supply chain integration plays a partial mediating role in the positive impact of enterprise digital transformation on supply chain resilience, and enterprises can promote digital transformation and enhance supply chain resilience through this mechanism of supply chain integration. The moderating effect test results show that both environmental uncertainty and enterprise risk-taking level play a positive moderating role in the relationship between digital transformation and supply chain resilience. This study can provide ideas and lessons for industrial supply chain management within the age of digitization.

1. Introduction

Given the current situation of globalization in the economy, with further refinement of the international division of labor, the global supply chain network continues to expand in geographic space, and its structure is becoming more and more complex, which makes it more vulnerable to the influence of external environmental factors [1]. At the same time, the trends of trade protectionism and anti-economic globalization are intensifying along with the intensification of geopolitical conflicts, and the global governance system is undergoing profound changes. Uncertainties, instabilities, and unpredictable events have significantly increased as a result of this set of adjustments. Whether by natural disasters, extreme weather, or wars, global supply chains can be hit, further highlighting their vulnerability and risk of disruption. In such a crisis environment, the importance of supply chain resilience has gradually been noticed, with the potential to provide strong response, coping, and adaptive capabilities to help enterprises handle difficulties or even turn the crisis into an opportunity [2]. Meanwhile, supply chain resilience plays a key role in enterprises’ pursuit of long-term development goals. Therefore, enhancing supply chain resilience has become a common focus of attention in both theory and practice.

Businesses are always undergoing digital transformation due to the quick growth of digital technologies. Artificial intelligence, cloud computing, and big data are all advancing quickly, and the full impact of the digital economic era is drawing near. Digital technology has emerged as a key engine of long-term economic expansion [3]. The China Academy of Information and Communications Technology’s “China Digital Economy Development Report (2024)” notes that the country’s digital economy grew by a noteworthy 7.4% in 2023. With a total value of around CNY 53.9 trillion, or 42.8% of GDP, the digital economy plays a crucial role in China’s economic structure [4]. In this process, digital technology, smart manufacturing, and the construction of information technology systems have become key factors in the transformation of the supply chain. The improvement of supply chain resilience and the digital transformation of businesses are closely related [5]. Given that supply chains are constantly changing and developing, it is of great value to study this change over time.

At present, enterprises are confronted with an environment of complex and changeable risks and accelerated technological innovation. Enhancing the resilience of the supply chain has become a key approach for modern manufacturing to maintain a competitive edge and achieve long-term sustainable development [6]. Digital technology can drive enterprises to optimize processes, enhance efficiency, reduce costs, promote innovation, and ensure product quality, among other aspects [7]. It not only provides technical support for the transformation and upgrading of enterprises, but also becomes an important driving force for promoting the sustainable and high-quality development of industries and the economy [8]. With the continuous deepening of global industrial transformation, the supply chains of manufacturing enterprises have gradually become unstable and complex. These characteristics have exacerbated the complexity of manufacturing operations and also brought greater challenges. Therefore, building a resilient and sustainable supply chain system through digital transformation has become an important path for the development of manufacturing enterprises.

The research significance of this article is as follows:

Theoretical significance: This study explores the key paths and mechanisms by which enterprises enhance the resilience of their supply chains through digital transformation in the current complex and volatile business environment. On the one hand, it enriches the relevant research on the influencing factors of supply chain resilience; on the other hand, it can also enrich the research on the effects of enterprise digital transformation.

Empirical significance: This work incorporates supply chain integration into the research framework to explore the impact of supply chain integration as a mediating variable on the role of enterprise digital transformation in enhancing supply chain resilience. Further, the moderating role of environmental uncertainty and the level of enterprise risk taking in the relationship between the two are considered. This provides valuable inspiration for enterprises on how to effectively enhance supply chain resilience through digital transformation under different environmental conditions.

Practical significance: This article links enterprise digitalization with the transformation of supply chain resilience, which is conducive to helping enterprises master methods to enhance supply chain resilience. Secondly, at present, the digital economy has taken a central position in economic development. The government has adopted a series of measures to encourage enterprises to implement digital transformation to promote the development of the digital economy and better adapt to the new driving forces of economic development. We conduct an in-depth analysis of the correlation between the current national policy environment and the actual demands of enterprises to provide systematic strategic guidance for them. This can help enterprises enhance their competitiveness and adaptability through digital transformation to strengthen supply chain resilience and promote sustainable development.

Driven by digital technologies, the digital trend is profoundly affecting all industries, not only changing the organizational structure of enterprises, but also reshaping their production and business models, which represents the full arrival of the digital economy era [9,10]. However, with the increasing uncertainty and crisis events faced by enterprises, the environment is growing increasingly complicated and changeable. Helping enterprises cope with challenges and achieve steady growth through digital transformation has become an urgent problem to be solved.

The influence mechanism between digital transformation and supply chain resilience has become a key topic of discussion in the academic community. In the academic discussions in the field of supply chain resilience, the existing literature focuses on theoretical elaboration and qualitative research [11]. In contrast, the application of quantitative research methods is relatively limited. In the existing literature, there are still relatively few studies that regard enterprise digital transformation as an independent variable and supply chain resilience as a dependent variable. Some academic studies have explored the initial theoretical framework and operational mechanism of the relationship between enterprise digital transformation and supply chain resilience. However, the current research field still has room for deepening and expansion, and further empirical analysis is urgently needed to enrich the knowledge system in this field.

• There is also a lack of research based on a systematic theoretical framework in the existing literature, especially in the field of antecedents for forming supply chain resilience, where the influence mechanism of enterprise digital transformation on this process has not been comprehensively explored.
• Most of the existing studies focus on analyzing the immediate effects of digital transformation on supply chain resilience, rather than delving into the relationship between the two in a long-term context. Supply chain resilience plays a crucial role in ensuring the sustainable development of enterprises.
• At present, most of the relevant literature focuses on case analyses or small-sample questionnaire surveys. Articles that adopt large-sample empirical analyses are relatively scarce, which limits the general applicability and stability of the research results.

To address the research gap mentioned above, the following are the innovation points and research directions of this paper:

• Empirical analysis methods are used to explore whether the digital transformation of enterprises can positively affect the resilience of the supply chain.
• To further examine the mechanism of action between the two, supply chain integration is introduced as a mediating variable, deepening the understanding of the decisive elements for enhancing supply chain resilience and expanding the economic effect scope of enterprise digital transformation.
• Environmental uncertainty and enterprise risk taking are introduced as moderating variables to explore their mechanism of action in the correlation between enterprise digital transformation and supply chain resilience. This innovation point helps to deeply explore the influence mechanism of environmental uncertainty and the level of enterprise risk taking on supply chain resilience, providing a reference for formulating effective strategies to deal with the complex and changeable business environment.

2. Theoretical Analysis and Research Hypotheses

2.1. Theoretical Basis

The resource-based theory holds that the competitive advantage of an enterprise stems from its unique and difficult-to-imitate resources and capabilities. Digital transformation, by integrating data, technology, and digital infrastructure, can enhance an enterprise’s dynamic capabilities and enable it to respond more flexibly to the risk of supply chain disruptions. Digital technologies such as the Internet of Things and artificial intelligence can optimize inventory management and enhance the accuracy of demand forecasting, enabling enterprises to adjust resource allocation more quickly in the face of external shocks and thereby improving the resilience of the supply chain.

The dynamic capability theory further emphasizes the ability of enterprises to integrate, construct, and reorganize internal and external resources in a rapidly changing environment. Digital transformation, through real-time data analysis and intelligent decision support systems, endows enterprises with stronger environmental perception capabilities and adaptability, enabling them to quickly identify supply chain risks and take countermeasures, such as achieving supplier diversification through digital supply chain platforms and reducing the risk of single dependence.

Information processing theory points out that organizations need to deal with environmental uncertainties through effective information processing. Digital transformation, through technologies such as big data analysis and cloud computing, significantly enhances the speed of information acquisition, processing, and response for enterprises, making supply chain management more transparent and collaborative. This reduces information delay and distortion and strengthens the buffering capacity against unexpected events.

The social network theory holds that enterprises are embedded in complex supply chain networks, and their resilience depends on the information flow and collaboration efficiency within the network. Digital technologies such as blockchain and smart contracts can enhance trust and cooperation among supply chain partners, reduce transaction costs, and facilitate rapid information sharing and collaborative recovery during crises, thereby improving the overall network’s risk resistance capacity.

These theories jointly support the theoretical assumption that digital transformation positively influences the resilience of an enterprise’s supply chain by optimizing resource utilization, enhancing dynamic adaptability, improving information processing efficiency, and strengthening supply chain collaboration.

2.2. Digital Transformation and Supply Chain Resilience

Digital transformation of enterprises refers to the improvement of management efficiency and enhancement of innovation capability by integrating modern information technology and digital tools into its various business segments and operation modes [12,13,14]. Digital transformation covers technologies such as big data analytics, the Internet of Things, and blockchain, which allow enterprises to realize the intelligence and collaboration of business processes and drive innovation in management models. It was not until the late 1980s that researchers in business and management began to focus on the concept of “resilience” [15]. Supply chain resilience refers to the ability of a supply chain to minimize interruption risks (such as internal operational failures or external natural disasters), control their impact, and quickly recover through adaptive strategies, as well as restore or even improve operational robustness through rapid response mechanisms and structural control [16].

With the help of big data, businesses undergoing digital transformation could gather and evaluate data from every part of the supply chain in real time, providing comprehensive visual information and predictive analysis [17]. Based on these data, businesses can do precise demand forecasting, inventory control, and logistics scheduling to increase the supply chain’s adaptability and responsiveness [18]. The transparency of information through digital technology enables enterprises to detect anomalies in the supply chain in a timely manner and take swift countermeasures to reduce the risk of supply chain disruptions. Through process automation and intelligence, digital transformation increases the supply chain’s operational efficiency [19] and reduces waste and costs. For example, through the intelligent logistics system, enterprises can optimize distribution paths, reduce transportation costs, and improve logistics efficiency. Digital technology empowers the supply chain to flexibly respond to changes in market demand and unexpected events. Enterprises can build intelligent manufacturing systems to realize the rapid adjustment of production lines and flexible production according to changes in the market [20]. The application of digital platforms enables enterprises to respond quickly to customer needs and provide personalized and customized services [21], enhancing customer satisfaction and market competitiveness. Alibaba has realized the digital management of the whole process of logistics by building an intelligent logistics network. Using artificial intelligence technology, the Cainiao network has optimized the warehouse layout and distribution paths, significantly improving logistics efficiency and service quality [22]. Digital transformation allows enterprises to develop supply chain finance by utilizing blockchain technology, thus enhancing supply chain transparency and traceability and reducing credit risk and operational risk [23]. Digital transformation helps improve supply chain resilience, adaptability, and recovery, which in turn strengthens the supply chain’s resilience. Therefore, we propose the first hypothesis:

H1. 

Digital transformation can significantly improve manufacturing supply chain resilience.

2.3. The Mediating Role of Supply Chain Integration

Supply chain integration can be divided into internal integration and external integration. Under the role of internal integration, each link and each department within the enterprise fully coordinates and cooperates with each other, systematically carries out data sharing and process optimization, and improves the operational efficiency within the enterprise. Through external integration, enterprises are able to strengthen synergies with upstream and downstream partners in the supply chain and promote the realization of comprehensive optimization of the entire supply chain with the help of strategic partnerships and the connection of information systems [24,25,26].

Digital transformation has invested in building information systems to enable information sharing and transparency in all segments of the supply chain, promoting integration between upstream and downstream enterprises in the supply chain. Digital transformation improves the efficiency of resource allocation and optimizes the operation flow of each link in the supply chain through process automation and intelligence. Enterprises can use IoT technology to realize intelligent monitoring and scheduling of production equipment and logistics facilities, improve the efficiency of production and logistics, and reduce operating costs [27]. Digital transformation is conducive to resource integration between upstream, midstream, and downstream enterprises in the supply chain, strengthening collaboration as well as relationship stability and enhancing the overall coordination and synergy of the supply chain [28]. Enterprises can realize the rapid coordination and resource scheduling of each link in the supply chain and improve the response speed and flexibility in dealing with emergencies. Supply chain collaboration platforms enabled by digital technologies allow enterprises to realize real-time communication and collaboration with upstream suppliers and downstream customers, enhancing supply chain resilience.

The internal and external integration of the supply chain improves the operational efficiency and cost control of enterprises, enhances their ability to cope with market fluctuations, and promotes the improvement of supply chain resilience. Through data analysis and decision support systems for each link, supply chain integration helps the chain to improve its overall responsiveness and flexibility, respond swiftly to changes in the market and customer demands, and promptly modify its strategies and modes of operation [29]. Through the ERP system and SCM system, enterprises can monitor the inventory level and production progress in real time and adjust the production and supply plan in a timely manner [30]. Supply chain resource integration promotes the visibility and visualization of enterprise supply chain processes, enabling enterprises to identify and solve potential problems in time. Through data analysis and predictive modeling, enterprises can identify supply chain risks in advance, formulate effective risk management plans and response strategies [25,31,32]. Supply chain integration enables enterprises to restore production and supply chain operations faster and achieve rapid recovery and reconstruction by coordinating the resources and capabilities of each link in the supply chain. Therefore, this study proposes the following hypothesis:

H2. 

Supply chain integration plays a mediating role in the impact of enterprises’ digital transformation on manufacturing supply chain resilience.

2.4. The Moderating Role of Environmental Uncertainty and Corporate Risk-Taking

This study bases its analysis on two perspectives: environmental uncertainty outside the organization and corporate risk taking within the organization, in order to reveal the regulating mechanism of the two in the impact of the enterprise’s digital transformation on the resilience of the supply chain.

Environmental uncertainty in this article refers to the uncertainty of the social and technological environment in which a company operates, namely the economic, political, and social environment. From an organization’s external perspective, being in an environment of uncertainty represents the instability of expectations, which indicates the scarcity of resources and the intensity of competition in the marketplace. Enterprises need to obtain resources from internal and external environments for value creation. As an open system, strategy and decision planning are affected by internal and external environments [33]. Changes in environmental uncertainty can also have an impact on planning for digital transformation. When enterprises are placed in a highly uncertain environment, they are more likely to invest in digital activities for competitive advantages in order to improve the resilience of their supply chains. At the micro level, environmental uncertainty may enhance managers’ sensitivity to decision-making information and facilitate the acquisition and integration of corporate resources, thus affecting the decision-making and transformation effects of corporate digital transformation [34]. An increase in economic policy uncertainty may promote corporate innovation and drive enterprises to invest funds in digital transformation projects [35], which can help corporations improve supply chain resilience. Digital transformation is conducive to the flexibility of enterprises when facing changes in internal and external environments, maintaining the stability and development of enterprises in a high-uncertainty environment, and strengthening the connection with upstream and downstream enterprises in the supply chain, thus contributing to the enhancement of the effect of enterprise digital transformation [36,37]. Facing high environmental uncertainty, enterprises can access and internalize innovative resources and knowledge through digital transformation. Seeking support from external sources and interacting with external knowledge and expertise increases supply chain resilience. Therefore, this study proposes the following hypothesis:

H3a. 

Environmental uncertainty positively moderates the relationship between enterprises’ digital transformation and supply chain resilience.

From an intra-organizational perspective, enterprises with high levels of corporate risk taking may be more inclined to choose digital transformation as an option when it comes to improving supply chain resilience. In the process of making decisions, enterprises need to comprehensively consider the variability of the market environment, the complexity of the competitive landscape, the uncertainty of technological development, and the actual situation of their own management capabilities, and therefore must make a trade-off between the risks they may face and the expected benefits [38]. Digital transformation not only requires long-term as well as large investments, but also has a long payback cycle, and enterprises with a high level of corporate risk taking have a relatively small short-term financial burden and are more willing to invest in innovation as well as digitization activities [12]. Financial pressures and revenue objectives of enterprises with high corporate risk-taking levels become one of the factors affecting the effectiveness of digitization on supply chain resilience enhancement. In addition, managers of enterprises with high corporate risk-taking levels are more tolerant of risk and uncertainty in corporate innovation projects and more optimistic about the expected benefits of digital transformation [39,40], which in turn increases digital investment and innovation funding. Digital transformation also requires enterprises to make adjustments to their internal culture and optimize and upgrade their organizational structure to meet business needs and technological development in the digital era, and at the same time, they need to strengthen the cultivation and skill enhancement of digital talents [41], which requires a large amount of human resource investment and a long-term talent cultivation plan. Enterprises with a high level of corporate risk taking are more motivated and capable of advancing the digital transformation process as they value long-term business outcomes and profits more. Therefore, this study proposes the following hypothesis:

H3b. 

Corporate risk-taking level positively moderates the relationship between enterprises’ digital transformation and manufacturing supply chain resilience.

The conceptual model constructed based on the analyzed content of the above theoretical mechanisms is shown in Figure 1.

3. Research Design

3.1. Research Method

Based on the panel data of A-share manufacturing listed companies in China from 2012 to 2022, a fixed-effect regression model was constructed to quantitatively analyze the correlation effect between supply chain resilience and the degree of enterprise digital transformation, and robustness tests for multiple methods were conducted. The mediating effect of supply chain integration and the moderating effect of environmental uncertainty and the level of enterprise risk assumption were further examined, ultimately providing reliable empirical evidence support for the digital transformation decision-making and supply chain risk management of listed companies.

The fixed effect model is an econometric method used for panel data analysis. Its core idea is to control unobservable heterogeneity that does not change over time by introducing fixed effects at the individual or time level, thereby alleviating the problem of bias in omitted variables. This model sets dummy variables for each individual or time point or eliminates the time-invariant features among individuals in the intra-group deviation transformation, so that the estimation results only reflect the net influence of the explanatory variables fluctuating over time within the individuals.

3.2. Sample Selection and Data Sources

This study selects Chinese A-share listed manufacturing enterprises from 2012 to 2022 as the research object, and the data are processed as follows: (1) excluding samples with the existence of ST (including *ST), a significant amount of missing relevant data, and the insolvency of samples; (ST refers to a special treatment for a company that has been operating at a loss for two consecutive years; ST * refers to a company operating at a loss for three consecutive years and receiving a delisting warning) and (2) shrink-tailing the variables at the level of 1 percent and 99 percent in order to eliminate the influence of outliers. After the above data processing, a total of 10,729 sample observations were obtained. The financial data used in this study come from the Cathay Pacific database and the iFinD database, and the internal control index comes from the Dib database.

3.3. Variable Definition and Measurement

• Explained variable

Supply chain resilience ($Resil$). Based on the research of Kuiran Shi et al. (2024) [42], the enterprise supply chain resilience is measured from two dimensions: fracture resilience and impact resilience, where the former is measured from the dimension of robustness and liquidity, which embodies the ability to cope with and adjust to external shocks it encounters, and the latter is measured from the dimension of vulnerability and development, which embodies the ability to withstand shocks and sustain development. According to the above four dimensions, the corresponding indicators are screened, and the specific indicator system is shown in

Table 1. Indicator system for supply chain resilience.

Target Layer	First-Level Indicators (Criterion Layer)	Second-Level Indicators (Element Layer)	Third-Level Indicators (Indicator Layer)	Description of Indicators
Supply Chain Resilience	Fracture Resilence	Robustness	Operating Income	Gross Operating Income (billion)
			Asset–liability Ritio	Total Liability/Total Asset (%)
			Total Profit	Operating Profit (billion)
		Liquidty	Capital Flow	Operating Expense (billion)
			Technical Flow	R&D Expense (billion)
	Impact Resilence	Vulnerability	Current Ratio	Current Asset/Current Liability (%)
			Operating Profit Ratio	Operating Profit/Operating Income (%)
			Inventory Turnover	Operating Cost/Average Inventory (%)
		Development	Non-current Asset	Non-current Asset (billion)
			Investment Activity	Cash Flow from Investing Activity (billion)
			Financing Activity	Cash Flows from Financing Activity (billion)

. The entropy weight method is used to measure the supply chain resilience level of each enterprise.

2.

Explanatory variable

Digital transformation ($Dig$). The text analysis method proposed by Fei Wu et al. (2021) [43] for measuring the level of digital transformation of enterprises is significantly reasonable. This method is based on the text data of the annual reports of listed companies. Firstly, it can objectively capture the substantive content of the implementation of the enterprise’s digital strategy and avoid the subjective bias of traditional questionnaire surveys. Secondly, it reflects the true degree of an enterprise’s digital investment through massive unstructured text data, breaking through the limitations of financial indicators. Finally, it can meet the requirements of repeatability and large sample coverage in academic research.

The method of text analysis takes the annual reports of listed companies as the object of text analysis, and the annual reports of listed companies can be obtained from the Cathay Pacific database. A keyword library is built for digital text analysis and Python 3.12 crawler technology is used to extract the words matching the keyword library in the enterprise annual report. Then, the occurrence frequency of these words is summarized. Finally, in order to eliminate the typical “right-bias” feature of this type of data, the frequency of word occurrence is increased by 1, and then the logarithm is taken to measure the level of digital transformation. The keyword libraries used are listed in the Appendix A.1.

3.

Mediator variable

Supply chain integration ($SCI$). There is a close relationship between the level of supply chain integration of an enterprise and the scale and stability of the business of the enterprise and its partners. When the business scale of both partners is large and stable, enterprises are more likely to take proactive measures to deepen supply chain integration, with the aim of optimizing resource allocation and strengthening the efficiency of cooperation. Conversely, if the business is smaller or more volatile, enterprises may reduce the resources invested in supply chain integration. Based on the study of Kehai Sheng (2022) [44], customer integration is measured by the indicator of the share of total annual purchases of the enterprise’s top five customers, while supplier integration is measured by the share of total purchases of the enterprise’s top five suppliers. Supply chain integration involves not only external integration between enterprises and upstream and downstream enterprises, but also the management of enterprises’ internal operations, i.e., internal integration, which is measured by the internal control index. Therefore, a comprehensive indicator system is constructed and measured using the entropy method.

4.

Moderating variables

Environmental uncertainty ($EU$). This study draws on the measures of Ghosh and Olsen (2008) [45] and Huihui Shen (2010) [46] and uses the industry-adjusted coefficient of variation in corporate sales revenue to calculate the environmental uncertainty faced by enterprises.

$$Sale=\beta +\gamma Year+\epsilon$$

(1)

Among them, $Sale$ is turnover, $Year$ is the annual variable, and $\epsilon$ represents the perturbation term. To calculate environmental uncertainty before filtering out industry factors, environmental uncertainty is calculated as the ratio of the standard deviation to the median of the most recent five-year balance sheet. Environmental uncertainty ($EU$) is calculated by dividing environmental uncertainty before filtering out the industry factor by the industry median for the same indicator in the same year to filter out the industry factor.

Corporate risk taking ($Risk$). With reference to Faccio (2016) [47] and others, corporate risk taking is assessed using the degree of volatility of the enterprise’s return on assets ($Roa$) over the observation period. $Roa$, the ratio of $EBIT$ (earnings before interest and taxes) to year-end total assets, is one of the important financial indicators of an enterprise. In order to reduce the impact of industry and economic fluctuations and accurately reflect the individual risk characteristics of enterprises, $Roa$ is subtracted from the annual industry average to obtain $Adj\_Roa$(2). Using every three years as an observation period, the rolling standard deviation of $Adj\_Roa$ (3) is calculated and a risk indicator is computed to measure corporate risk taking.

$${Adj\_Roa}_{i,n}={\displaystyle \frac{{EBIT}_{i,n}}{{ASSET}_{i,n}}}-{\displaystyle \frac{1}{X}}\sum _{k=1}^X{\displaystyle \frac{{EBIT}_{i,n}}{{ASSET}_{i,n}}}$$

(2)

$${Risk}_{i,n}=\sqrt{{\displaystyle \frac{1}{T-1}}\sum _{n=t-2}^t{\left({Adj\_Roa}_{i,n}-{\displaystyle \frac{1}{T}}\sum _{n=t-2}^t{Adj\_Roa}_{i,n}\right)}^2}|\mathrm{T}=3$$

(3)

5.

Control variables

Referring to the study of Shushan Zhang et al. (2023) [48], the control variables including enterprise size ($Size$), cash flow ($Cashflow$), enterprise age ($Listage$), board size ($Board$), proportion of independent directors ($Indpe$), enterprise growth ($Growth$) and return on equity ($ROE$) are selected to reflect the enterprise’s financial and growth factors as well as corporate governance aspects of the control variables. The description of the variables is shown in

Table 2. Variable definitions and measures.

Variable Type	Variable Name	Variable Symbol	Variable Measurement
Explained variable	Supply Chain Resilience	Resil	Obtained by the index system constructed in Table 1
Explanatory variable	Digital Transformation	Dig	ln(1 + total number of digitized keywords)
Mediator variable Supply Chain Integration (SCI)	Supplier Integration	SI	Top five supplies’ purchase share/total purchase share
	Customer Integration	CI	Top five customer sales share/total sales
	Internal integration	LNICI	Internal control index
Moderating variable	Environmental Uncertainty	EU	Coefficient of variation in abnormal income in the past five years after industry adjustment
	Corporate Risk-Taking	Risk	Degree of fluctuation in enterprises’ Roa over the observation period
Control variable	Enterprise Size	Size	Natural logarithm of total assets for the year
	Enterprise Age	ListAge	ln(current year − listing year + 1)
	Return on Equity	ROE	Net profit/average balance of shareholders’ equity
	Enterprise Growth	Growth	Current year’s operating income/previous year’s operating income − 1
	Board Size	Board	Natural logarithm of the number of board members
	Proportion of Independent Directors	Indep	Independent directors divided by number of directors
	Cash Flow	Cashflow	Net cash flows from operating activities divided by total assets

.

3.4. Model Construction

According to the previous hypothesis analysis, the following regression model is constructed to test H1:

$${Resil}_{i,t}={\propto }_0+{\beta }_0{Dig}_{i,t}+\sum \gamma {Controls}_{i,t}+{\delta }_i+{\delta }_t+{\delta }_j+{\epsilon }_{i,t}$$

(4)

In order to further test the mediating role of supply chain integration, drawing on the method of Zhonglin Wen et al. (2004) [49] to study the mediating effect, Models (5) and (6) are constructed based on Model (4) for testing the mediating effect of supply chain integration in the relationship between the two, i.e., testing hypothesis H2.

$${SCI}_{i,t}={\propto }_1+{\beta }_1{Dig}_{i,t}+\sum \gamma {Controls}_{i,t}+{\delta }_i+{\delta }_t+{\delta }_j+{\epsilon }_{i,t}$$

(5)

$${Resil}_{i,t}={\propto }_2+{\beta }_2{Dig}_{i,t}+{\mathsf{\beta }}_3{SCI}_{i,t}+\sum \gamma {Controls}_{i,t}+{\delta }_i+{\delta }_t+{\delta }_j+{\epsilon }_{i,t}$$

(6)

Models (7) and (8) are used to test the moderating effects of environmental uncertainty and corporate risk taking, i.e., hypotheses H3a and H3b.

$${Resil}_{i,t}={\phi }_0+{\phi }_1{Dig}_{i,t}+{\phi }_2{EU}_{i,t}+{\phi }_3{Dig}_{i,t}\times {EU}_{i,t}+\sum \gamma {Controls}_{i,t}+{\delta }_i+{\delta }_t+{\delta }_j{+\epsilon }_{i,t}$$

(7)

$${Resil}_{i,t}={\mu }_0+{\mu }_1{Dig}_{i,t}+{\mu }_2{Risk}_{i,t}+{\mu }_3{Dig}_{i,t}\times {Risk}_{i,t}+\sum \gamma {Controls}_{i,t}+{\delta }_i+{\delta }_t+{\delta }_j+{\epsilon }_{i,t}$$

(8)

Among them, $i$ denotes an individual enterprise and $t$ denotes a year; $Resil$ is supply chain resilience, $Dig$ is the level of enterprise digital transformation, $SCI$ is the mediating variable supply chain integration, $EU$ is environmental uncertainty, $Dig\times EU$ is the cross-multiplier of enterprise digital transformation and environmental uncertainty, $Risk$ is enterprise risk taking, $Dig\times Risk$ is the cross-multiplier of enterprise digital transformation and corporate risk taking; $Controls$ represents a series of enterprise-level control variables, the firm-fixed effect ${\delta }_i$ controls individual differences, the year-fixed effect ${\delta }_t$ controls time differences, the industry-fixed effect ${\delta }_j$ controls industry differences, and $\epsilon$ represents the perturbation term.

4. Empirical Results and Analysis

4.1. Descriptive Statistics

A total of 10,729 samples were obtained through data collection and organization, and the descriptive statistics are shown in

Table 3. Descriptive statistics.

Variables	Sample Size	Standard Deviation	Minimum	Maximum	Average
Resil	10,729	0.0407	0.0204	0.4435	0.0540
Dig	10,729	1.2193	0.0000	5.0304	1.4019
SCI	10,729	0.1237	0.1170	0.7649	0.3421
EU	10,729	0.8101	0.1412	6.7926	1.1412
Risk	10,729	0.0307	0.0012	0.2676	0.0292
Size	10,729	0.9927	19.9412	25.7906	21.9948
ROE	10,729	0.1092	−0.6922	0.4151	0.0685
Cashflow	10,729	0.0624	−0.1618	0.2656	0.0532
Growth	10,729	0.2953	−0.4912	2.4450	0.1574
Board	10,729	0.1857	1.6094	2.6391	2.0990
Indep	10,729	0.0532	0.3125	0.5714	0.3767
ListAge	10,729	0.7034	0.6931	3.3673	2.0308

. The minimum value of the supply chain resilience indicator ($Resil$) is 0.0204 and the maximum value is 0.4435, indicating that there is significant variability in supply chain resilience across enterprises. The standard deviation of the Degree of Digital Transformation ($Dig$) is 1.2193, with minimum and maximum values of 0 and 5.0304, respectively, which also indicates that the degree of digital transformation varies widely among enterprises and that there are still enterprises that have not yet begun to implement digital transformation. All key variables are within reasonable limits.

4.2. Basic Regression Analysis

The basic regression Model (4) is used to study the direct impact of enterprise digital transformation on manufacturing supply chain resilience, and the results are shown in

Table 4. Basic regression results.

Variables	(1)	(2)
	Resil	Resil
Dig	0.0017 *** (6.3622)	0.0005 ** (2.2128)
Size		0.0245 *** (45.4506)
ROE		0.0029 (1.5536)
Cashflow		0.0179 *** (5.7537)
Growth		−0.0025 *** (−4.3027)
Board		0.0015 (0.7968)
Indep		0.0198 *** (3.3173)
ListAge		−0.0127 *** (−14.6701)
_cons	0.0516 *** (127.6715)	−0.4724 *** (−37.3439)
Time/enterprise/industry fixed	YES	YES
N	10,729	10,729
adj. R2	0.853	0.882

Note: **, *** indicate significant at the 5% and 1% levels of statistical significance, respectively. Values in parentheses are t-values (The same below).

. Column (1) presents the results of the baseline regression with only the key explanatory variables, while control variables are added in column (2), and enterprise, industry, and year-fixed effects are added in both columns (1) and (2). The results of the analysis show that the coefficient on enterprises’ digital transformation is significantly positive, indicating that enterprises’ digital transformation has a significant positive impact on manufacturing supply chain resilience. H1 is demonstrated.

4.3. Robustness Test

• Lagged variable

Because of the possible causal association between digital transformation and supply chain resilience, to eliminate this effect, this study adopts digital transformation lagged by one period as a new explanatory variable, i.e., the supply chain resilience data of the previous period is used to regress with digital transformation of the current period. The results are shown in Model (1) and (2) in

Table 5. Robustness test results I.

Variables	One Period Behind		Substitution of Variables
	(1)	(2)	(3)	(4)
	Resil	Resil	Scr	Scr
L.Dig	0.0015 *** (5.2252)	0.0006 ** (2.3519)
DT			0.0502 *** (8.0022)	0.0071 *** (2.6566)
_cons	0.0516 *** (124.7422)	−0.4999 *** (−33.5382)	1.9595 *** (99.9454)	2.5454 *** (23.9290)
controls	NO	YES	NO	YES
Time/enterprise/industry fixed	YES	YES	YES	YES
N	7859	7859	10,962	10,962
adj. R2	0.852	0.882	0.500	0.912

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

, and the results are consistent with the previous section.

2.

Replacement variable

Replacement of the measurement methods of the explained variables and explanatory variables in the previous study is conducted by adopting the study of Chenyu Zhao et al. (2021) [50], specifically also using text analysis to conduct word frequency statistics on corporate annual reports, but with a different selected keyword library, replacing the digital transformation (DT) indicator. Referring to the study of Huiyan Wang et al. (2023) [51], factor analysis is used to construct a supply chain resilience indicator (Scr). The results are shown in Model (3) and (4) in Table 5, and the conclusions remain consistent with the previous study.

3.

Instrumental variables (IV-2SLS)

In empirical analysis, the instrumental variable method (IV) is mainly used to solve endogeneity problems. Endogeneity is the correlation between explanatory variables and error terms, which leads to biases in regression analysis. Instrumental variables are those that are highly correlated with endogenous explanatory variables but do not directly affect the dependent variable. Endogeneity bias can be eliminated for instrumental variables through the two-stage least square method (2SLS). The operation steps are as follows: In the first stage, the endogenous explanatory variable is regressed against the instrumental variable and other exogenous control variables, and then the predicted value of the endogenous variable is obtained. This step strips away the interfering parts related to the error term. In the second stage, the original dependent variable is regressed against the predicted values obtained in the first stage and other control variables. At this point, since the predicted values are generated by the instrumental variables and satisfy exogeneity, the final regression can reflect a robust causal effect.

To eliminate the problem of possible reverse causality in the results of the benchmark regression, drawing on the prior literature, landline telephone data from 1984 are used as an instrumental variable to measure the level of digital transformation of manufacturing enterprises. Given the cross-sectional nature of historical data, time series variables are introduced to construct interaction terms with the historical landline data for regression analysis. Specifically, the natural logarithm of the number of internet access points in the province where the enterprise is located in the previous year is calculated and multiplied by the number of landline telephones per 10,000 people in the area in 1984 to form a new instrumental variable. According to the results in

Table 6. Robustness test results II.

Variables	(1)	(2)	(3)	(4)
	Ininv_num	Ininv_quote
	Phase I	Phase II	Phase I	Phase II
IV	0.1885 *** (17.36)		0.0001 *** (17.21)
Dig		0.0046 ** (3.30)		0.0048 *** (3.39)
controls	YES	YES	YES	YES
Time/enterprise/industry fixed	YES	YES	YES	YES
Cragg-Donald Wald F statistic	301.262		296.082
Anderson canon. corr. LM statistic	293.619 ***		288.706 ***
N	10,729	10,729	10,729	10,729

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

, the Cragg–Donald Wald F-statistic exceeds the 10% threshold, showing that the model passes the test for weak instrumental variables, while the Anderson LM statistic is significant at the 1% level, indicating that the model passes the test for non-identifiability. The first-stage regression coefficients for the instrumental variables are significantly positive, indicating a positive relationship between historical landline penetration and the digitization process of manufacturing enterprises. In the second stage regression, the regression coefficients for enterprises’ digital transformation are also significantly positive, indicating that the main findings of this study remain robustly established after controlling for reverse causation issues.

4.4. Mediating Effect Test

• Mediating effect

The theoretical hypothesis section proposes the mediating role of supply chain integration between enterprises’ digital transformation and supply chain resilience and analyzes the path between the two. Drawing on the mediation effect model of Zhonglin Wen et al. (2004) [49], Model (5) and Model (6) were established on the basis of Model (4), and the corresponding regression results of the model are shown in

Table 7. Mediating effect test results.

Variables	(1)	(2)
	SCI	Resil
Dig	0.0026 *** (2.7094)	0.0005 ** (2.0812)
SCI		0.0123 *** (4.6524)
_cons	0.6863 *** (13.5571)	−0.4809 *** (−37.6689)
controls	YES	YES
Time/enterprise/industry fixed	YES	YES
N	10,729	10,729
adj. R2	0.795	0.882

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

. As can be seen in Table 7, in column (1), the enterprise’s digital transformation is significantly positive at the 1% level, indicating that the enterprise’s digital transformation can improve the level of supply chain integration, i.e., the higher the level of the enterprise’s digital transformation, the higher the level of supply chain integration; in column (2), enterprise digital transformation is significantly positive at the 5% level and supply chain integration is significantly positive at the 1% level, indicating that the higher the digital transformation index of the enterprise, the higher the level of supply chain integration and the higher the supply chain resilience. According to the analysis of the mediation effect principle, it shows that supply chain integration plays a partial mediating role between enterprise digital transformation and supply chain resilience, which validates hypothesis H₂ of this study.

2.

Mediating effect test based on the Sobel and Bootstrap methods

Sobel and Bootstrap methods were used to further robustly analyze the mediating role of supply chain integration between enterprises’ digital transformation and supply chain resilience. The results, as shown in

Table 8. Mediating effect test based on Sobel and Bootstrap methods.

Sobel Method			Bootstrap Method				Result
Sobel Coefficient	SobelZ Value	p-Value	Ind_eff Factor	p-Value	95% Lower Limit	95% Cap
0.001	5.081	0.000	0.001	0.000	greater than zero	greater than zero	statistically significant

, show that the mediating effect of supply chain integration is statistically significant, demonstrating the robustness of its mediating effect.

4.5. Moderating Effects Test

In order to further test the moderating effects of environmental uncertainty and the level of corporate risk taking on the relationship between digital transformation and supply chain resilience, environmental uncertainty and the level of corporate risk taking, as well as their interaction terms with the digital transformation of the enterprises, are introduced into the regression model. As shown in Model (2) of

Table 9. Moderating effect test results.

Variables	(1)	(2)	(3)	(4)
	Resil	Resil	Resil	Resil
Dig	0.0005 ** (2.2069)	−0.0000 (−0.0588)	0.0005 ** (2.1323)	0.0002 (0.7103)
EU	0.0005 ** (2.0447)	−0.0001 (−0.3474))
Dig × EU		0.0005 *** (2.5917)
Risk			0.0297 *** (4.8940)	0.0124 (1.3369)
Dig × Risk				0.0110 ** (2.4734)
_cons	−0.4721 *** (−37.3223)	−0.4715 *** (−37.2737)	−0.4798 *** (−37.7078)	−0.4794 *** (−37.6844)
controls	YES	YES	YES	YES
Time/enterprise/industry fixed	YES	YES	YES	YES
N	10,729	10,729	10,729	10,729
adj. R2	0.882	0.882	0.882	0.882

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

, the estimated coefficient of the interaction term between environmental uncertainty and enterprise digital transformation is significantly positive, which indicates that environmental uncertainty plays a positive moderating role in the positive impact of enterprise digital transformation on supply chain resilience, i.e., the higher the environmental uncertainty, the more it promotes the positive impact of enterprise digital transformation on supply chain resilience. As shown in Model (4) of Table 9, the regression coefficient of the interaction term between corporate risk taking and enterprise digital transformation is also significantly positive, which indicates that corporate risk taking also produces a positive moderating effect in the process of enterprise digital transformation for supply chain resilience enhancement, and the enhancement of corporate risk taking is conducive to the promotion of enterprise digital transformation for supply chain resilience. This suggests that both environmental uncertainty and corporate risk taking positively moderate the relationship between digital transformation and supply chain resilience.

In summary, environmental uncertainty and corporate risk-taking level have significant moderating effects on the relationship between enterprise digital transformation and manufacturing supply chain resilience. Specifically, when environmental uncertainty is stronger and corporate risk taking is higher, the role of the digital transformation of enterprises in enhancing the resilience of manufacturing supply chains is more significant, thus validating the correctness of hypotheses H3a and H3b.

4.6. Heterogeneity Analysis

• Heterogeneity test based on market competition intensity

Resources in the market are limited, and the ease with which an enterprise can obtain resources from the market can be characterized by the intensity of competition in the market of the industry in which it operates. The industry Herfindahl index (HHI) is used as a measure of the intensity of competition in the market for manufacturing enterprises. If the Herfindahl index is below the annual median, it is labeled 0 and classified as a low-competition market group, and if it is above the median, it is labeled 1 and classified as a high-competition market group, and then regression analyses are performed separately. Enterprises in high market competition face a more intense industry market environment, need to continuously improve the supply chain resilience to stabilize the market position, and thus will be more inclined to carry out digital transformation, with the help of digital technology to reconstruct the competitive advantage and improve the stability of the supply chain. On the other hand, enterprises in low-competitive markets face relatively less risk and may have less incentive to adopt digital transformation. Models (1) to (2) in

Table 10. Heterogeneity test results.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
	Resil	Resil	Resil	Resil	Resil	Resil
	High Market Competition	Low Market Competition	High-Tech Industries	Non-High-Tech Industries	Nationalized Business	Non-State Enterprise
Dig	0.0007 ** (2.1589)	0.0004 (0.9375)	0.0006 ** (2.1228)	0.0003 (0.5294)	0.0015 ** (2.4475)	0.0005 * (1.7699)
_cons	−0.4585 *** (−28.8048)	−0.4989 *** (−22.0866)	−0.4634 *** (−32.8369)	−0.5097 *** (−16.9186)	−0.4455 *** (−13.4732)	0.4762 *** (−34.3670)
controls	YES	YES	YES	YES	YES	YES
Time/enterprise/industry fixed	YES	YES	YES	YES	YES	YES
N	6880	3672	8684	2033	2329	8361
adj. R2	0.845	0.918	0.884	0.873	0.928	0.834

Note: *, **, *** indicate significant at the10%, 5% and 1% levels of statistical significance, respectively. Values in parentheses are t-values.

show that digital transformation significantly enhances supply chain resilience in a high-competition market environment, while the positive effect of digital transformation on supply chain resilience is not significant in a low-competition market environment. The regression results support a stronger positive impact of digital transformation on supply chain resilience in highly competitive markets.

2.

Heterogeneity test based on the nature of enterprise

In the high-tech industry, enterprises usually have more complete technical reserves and previous development and construction of digital infrastructure, resources which promote further digital transformation of enterprises, establishing a digital advantage and breaking through barriers. These enterprises rely heavily on digital technology in their day-to-day operations and have maintained their industry leadership through continuous technological innovation and application promotion. High-tech enterprises not only have advantages in digital infrastructure but also have accumulated rich experience in technology application and management innovation and are better able to adapt to and navigate the changes brought about by digital transformation. High-tech and non-high-tech enterprises were divided into two sample groups for the benchmark regression analysis, and the results are shown in Models (3) and (4) in Table 10. The impact of digital transformation on supply chain resilience is significantly positive in the sample of high-tech enterprises. High-tech enterprises have achieved a high degree of integration and collaborative operation of their supply chains through digital technology, reducing the risks and losses associated with supply chain disruptions. In the sample of non-high-tech enterprises, the positive effect of digital transformation on supply chain resilience is not significant. This is mainly due to the fact that non-high-tech enterprises are relatively lagging behind in terms of digital infrastructure and technology application and are unable to fully utilize the advantages of digital technology, resulting in a less significant effect of digital transformation than high-tech enterprises.

3.

Heterogeneity test based on property right

From a resource perspective, state-owned enterprises (SOEs) usually have an advantage over non-SOEs, with greater resource reserves and indirect government support. Therefore, in the process of enterprise change, this rich reserve of resources enables state-owned enterprises to invest more resources in the process of digital transformation, thus advancing the process of digital transformation. The stability of state-owned enterprises in terms of organizational structure and management system enables them to adapt to new business models and management styles more quickly and maintain continuity and stability in the face of digital transformation. Given these factors, SOEs are likely to be more resilient to supply chains during digital transformation than non-SOEs. Therefore, the sample was divided into two categories, state-owned and non-state-owned enterprises, and separate regression analyses were conducted for each group. The results are shown in models (5) to (6) in Table 10, where the promotion effect of digital transformation on supply chain resilience enhancement of state-owned enterprises is significantly established, while the degree of significance of non-state-owned enterprises is lower than that of state-owned enterprises. Therefore, the positive impact of digital transformation on the supply chain resilience of state-owned enterprises is more significant.

5. Conclusions

Based on the panel data of Shanghai and Shenzhen A-share listed manufacturing enterprises from 2012 to 2022, this study empirically analyzes the impact of enterprise digital transformation on supply chain resilience, including the direct effect, mediating effect, and moderating effect of the three aspects of the influence mechanism, and the results show the following:

(1) Enterprises’ digital transformation significantly contributes to the resilience of manufacturing supply chains, and the finding remains valid after a series of robustness tests. (2) The results of the mediation effect analysis show that supply chain integration plays a partially mediating role in the positive impact of enterprise digital transformation on manufacturing supply chain resilience, indicating that enterprise digital transformation can enhance supply chain resilience by promoting supply chain integration. (3) The results of the moderating effect test found that both environmental uncertainty and corporate risk taking play a positive moderating role in the positive relationship between corporate digital transformation and manufacturing supply chain resilience. (4) The heterogeneity results show that digital transformation mainly affects the supply chain resilience of enterprises with high market competition, state-owned enterprises, and high-tech enterprises, while the effect on the supply chain resilience of enterprises in low-market industries, non-state-owned enterprises, and non-high-tech enterprises is not significant.

Based on these conclusions, the following practical insights are offered:

(1) As the main body of digital transformation, Chinese enterprises can fully leverage the policy dividends of the state and systematically enhance the resilience of the supply chain through the following approaches. Firstly, they can utilize the special subsidies under the “Overall Layout for Digital China Construction” [52] to deploy an AI-driven Internet of Things demand forecasting and real-time monitoring system. Second, they can participate in the “Action Plan for the Innovative Development of Industrial Internet” [53], establish a digital twin platform for connection with core suppliers, and achieve end-to-end visibility. Third, they can adopt blockchain solutions supported by the Blockchain Application Pilot Project (2023) to create a tamper-proof supplier network. We further emphasize how provincial policies such as Zhejiang Province’s “Digital Transformation Voucher Program” have helped small and medium-sized enterprises gradually implement cloud-based inventory management. Meanwhile, state-owned enterprises can take advantage of the “Guiding Opinions on Digital Transformation of State-owned Enterprises” [54] to institutionalize cross-supplier stress testing and transform policy opportunities into operational capabilities.

(2) Supply chain integration is a key factor in the digital transformation of enterprises to improve supply chain resilience. Enterprises should strengthen the synergy of upstream and downstream partners in the supply chain and establish a close information sharing mechanism. They can promote the integration of upstream, midstream, and downstream resources in the supply chain, realize the transparency of information in each link of the supply chain, and improve the coordination ability and reaction speed of the supply chain as a whole. Digital technology can be used to optimize supply chain processes, save intermediate link costs, and improve operational efficiency. Enterprises can use big data analysis technology to make accurate predictions and optimize scheduling for each link in the supply chain, reducing resource waste and operating costs. Supply chain visualization management can be implemented to identify and solve potential problems in a timely manner through real-time monitoring and data analysis and enhance the resilience and risk resistance of the supply chain. Digital transformation can help enterprises establish a flexible emergency response mechanism. Through intelligent systems, enterprises can realize real-time monitoring and rapid response to abnormal situations in the supply chain, ensuring that the stability and continuity of the supply chain is maintained during emergencies.

(3) Enterprises should establish a comprehensive environmental monitoring and risk warning system to obtain timely information on market dynamics and changes in the internal and external environment of the enterprise. Through big data analytics, we are able to predict environmental trends and develop response plans to counteract the negative impacts of environmental uncertainty on enterprises and the supply chains in which they operate. Enterprises should establish a scientific risk management system and formulate clear risk management strategies and measures. Enterprises can disperse supply chain risks through supply chain diversification strategies. For example, multiple suppliers and logistics channels can be selected to avoid supply chain disruptions due to problems with a single supplier or channel. Enterprises should remain sensitive to new technologies and actively explore and apply them to promote continuous innovation in supply chain management. In the process of digital transformation, we focus on the promotion and popularization of technology applications, so that more employees and partners in the supply chain can benefit from the enhancement of digital technology and promote the level of supply chain resilience.