Digitalization and Industrial Chain Resilience: Evidence from Chinese Manufacturing Enterprises

Abstract

(1) Background. The rapid development of the digital economy provides a new perspective for enhancing industrial chain resilience. This study examines how manufacturing firms’ digitalization affects their industrial chain resilience, drawing on resource dependence and dynamic capability theories, and explores spillover effects on upstream and downstream enterprises. (2) Data and Methods. Using panel data from Chinese listed manufacturing firms (2011–2023), we employ ordinary least squares (OLS) models to analyze the relationship, its mechanisms, and heterogeneity. We further match firms with their suppliers and customers to identify spillover effects. (3) Results. Digitalization significantly improves resilience, particularly by enhancing supply–demand matching and competitive capabilities. Effects are stronger for small, labor-intensive, and high-environment, social and governance (ESG) firms. Bargaining power and governance capability are key channels. Spillover effects are heterogeneous, with a stronger impact on downstream customers. (4) Discussion. The positive impact of digitalization varies by firm characteristics, and spillovers differ across the chain. These findings offer precise insights and policy implications for leveraging digitalization to strengthen industrial chain resilience.

1. Introduction

The global industrial chain system is undergoing unprecedented challenges, ranging from trade frictions to the COVID-19, and from natural disasters to geopolitical conflicts [1,2]. These black swan events continuously disrupt corporate industrial chain operations, highlighting the increasing vulnerability of global industrial networks. Amid the rapid development of the digital economy, digitalization has emerged as a revolutionary force, permeating various aspects of corporate operations [3] and gradually becoming a critical tool for enhancing industrial chain resilience. However, the impact of corporate digitalization on industrial chain resilience is dual-edged and complex. The Complex Adaptive Systems theory [4] posits that while corporate digitalization can strengthen the industrial chain resilience through data sharing and collaborative operations [5], it can also exacerbate digital divides among enterprises, leading to coordination breakdowns and reduced risk response capabilities across the industrial chain [6]. This “technological empowerment-systemic risk” duality constitutes the core tension in their relationship. Therefore, it is imperative to further investigate the relationship between corporate digitalization and industrial chain resilience, as well as the externalities it generates within industrial chain networks.

Existing research on digitalization has extensively explored its impacts on global division of labor [7,8], urban industrial chain resilience [9], industrial chain concentration [10], corporate productivity [11], financing constraints [12], and innovation capabilities [13]. It has also revealed that digital technologies can improve enterprises’ access to capital by enhancing investor information transparency [14]. However, within the complex structure of economic networks, significant research gaps remain regarding the profound effects of corporate digitalization on their respective industries chain resilience. First, although industrial chain resilience has garnered increasing attention [15,16], studies predominantly focus on using macro data or data from all listed companies, which underestimates the impact on manufacturing enterprises. Second, the specific mechanisms and pathways through which digitalization enhance industrial chain resilience remain unclear, as existing research has yet to thoroughly elucidate how technological applications translate into industrial chain resilience improvements. Third, while corporate digitalization can bolster the robustness of upstream suppliers and the adaptability of downstream customers [5,17], the channels through which spillover effects contribute to industrial chain resilience are not well-defined. Addressing these gaps is crucial for a comprehensive understanding of the interplay between digitalization and industrial chain resilience.

Therefore, to address the limitations of existing research, we utilize an unbalanced panel dataset of Chinese manufacturing enterprises from 2011 to 2023. Employing Ordinary Least Squares (OLS) estimation, we quantify how corporate digital transformation impacts industrial chain resilience. Additionally, we elucidate the underlying mechanisms and spillover effects on upstream and downstream enterprises. Our findings indicate that corporate digitalization significantly enhances industrial chain resilience, and the results remain robust to alternative specifications and endogeneity tests. Furthermore, we examine the heterogeneity of these effects across enterprises of varying sizes, capital intensities, ESG characteristics, and ownership structures. Corporate digitalization primarily influences industrial chain resilience by strengthening enterprises’ bargaining power and governance capabilities. We further reveal that digitalization generates asymmetric spillover effects within the industrial chain. While the resilience of downstream customers improves significantly, we find no statistically significant impact on upstream suppliers. These insights contribute to a deeper understanding of the role of digitalization in fostering industrial chain resilience and its broader economic implications.

This study contributes to the literature on enterprise digitalization and industrial chain resilience in three main theoretical aspects. First, this paper extends industrial chain resilience research to the enterprise level by identifying digitalization as a key micro-foundation of resilience. Compared to studies on corporate digitalization, which predominantly analyze data from all listed enterprises, this research focuses specifically on manufacturing enterprises [11,12,18,19]. the manufacturing sector, which not only features the most complex upstream and downstream division of labor but also exhibits industrial chain resilience that is more directly influenced by inter-enterprise digital coordination [16,20]. By focusing on manufacturing enterprises, this study demonstrates that industrial chain resilience emerges from firms’ digital transformation processes. This finding enriches complex adaptive systems theory and provides a micro-level explanation for the formation of industrial chain resilience.

Secondly, this study constructs a comprehensive theoretical framework based on resource dependence theory and dynamic capabilities theory to explain the mechanisms through which digitalization enhances industrial chain resilience. While existing research has examined the impact of digitalization on global value chain restructuring and the stability of regional supply networks [15,21], significant research gaps remain in the discussion of micro-level mechanisms. The dual-path theoretical framework we developed—which integrates bargaining power and governance capabilities with their synergistic effects—clarifies both the internal and external channels linking digitalization and resilience, thereby addressing the “black box” problem in the existing literature. Theoretically, it emphasizes digitalization as a strategic capability that reshapes power structures and governance efficiency within industrial chains.

Third, this study introduces a directional spillover perspective into the analysis of industrial chain resilience and elucidates the economic mechanisms underlying the asymmetric spillovers of enterprise digitalization on resilience between upstream and downstream entities in the industrial chain. Although some studies have explored the relationship between corporate digitalization and the resilience of upstream and downstream enterprises within industrial chains [5,17], the pathways and channels through which such externality operates need further research [22]. To address this, we employ enterprises’ bargaining power relative to their upstream and downstream counterparts to explain these externalities. This result expands industrial chain network theory by emphasizing the role of power asymmetry and information transmission in shaping resilience spillovers.

Overall, this study deepens the theoretical understanding of how enterprise digitalization contributes to industrial chain resilience, provides a clear micro-mechanism framework, and offers a new perspective on resilience spillover effects within industrial chain networks.

2. Theoretical Framework

2.1. Literature Review

The research stream related to this study mainly focuses on how digital technologies reshape economic systems, and the literature can be reviewed from the following perspectives.

• ICT, global trade dynamics, and firms’ export competitiveness

As a general-purpose technology, information and communication technology (ICT) has profoundly transformed global trade patterns by reducing information search costs and improving the coordination of cross-border payments and logistics. At the theoretical level, heterogeneous firm trade models represented by Melitz (2003) provide a fundamental framework for understanding how digital technologies alter both the fixed and variable costs of firms’ participation in export markets [23]. Empirical studies further indicate that the diffusion of ICT infrastructure not only enhances a country’s overall export sophistication [24], but also significantly promotes the development of the global digital economy, including improvements in productivity, innovation, and trade performance [25].

• Digitalization, green transformation, and environmental sustainability

Digital technologies play a complex dual role in promoting green growth and addressing environmental challenges. On the one hand, digital technologies can significantly improve energy and resource efficiency and enable more precise environmental regulation and carbon emission monitoring [26,27]. On the other hand, the energy consumption of digital infrastructure itself, together with the potential expansion of production and consumption scales induced by digitalization, may intensify environmental pressures [28]. Some studies further analyze the specific mechanisms and boundary conditions through which digital technologies affect environmental performance, emphasizing that digitalization exhibits a pronounced “double-edged sword” effect [29].

• Institutional and industrial policies and industrial transformation

The institutional environment and targeted industrial policies constitute critical contextual factors shaping technology absorption and economic transformation. For example, studies on digital infrastructure policies such as the “Broadband China” strategy show that these policies not only directly stimulate firms’ digital investment [30], but also indirectly influence industrial agglomeration and supply chain efficiency by improving regional connectivity [31,32]. Meanwhile, research on institutional arrangements such as free trade zones demonstrates that by reducing institutional transaction costs and facilitating knowledge spillovers, these policies promote industrial upgrading and the transformation of trade patterns [33,34].

• Technology diffusion, human capital, and inclusive development

The economic benefits of digital technologies ultimately depend on the absorptive capacity of micro-level actors. A large body of literature shows that the distribution of digital dividends is highly uneven and closely associated with human capital levels, skill structures, and social capital [35]. For instance, studies on the digital divide reveal how inequalities in access, usage, and digital skills further exacerbate existing economic and social disparities [3]. Moreover, executives’ digital cognition and employees’ digital skill endowments constitute the core of firms’ digital absorptive capacity [36], directly affecting the efficiency with which technological investments are translated into productivity gains and innovation outcomes [37].

These strands of literature provide a solid, multidimensional, and in-depth foundation for investigating the relationship between enterprise digitalization and the construction of industrial chain resilience.

2.2. Theoretical Analysis

Existing literature consistently indicates that digitalization can reshape supply chain relationships through information transparency [38] and network diversification [39], while also exerting direct impacts on enterprise operations [40] and innovation efficiency [41]. However, these discussions have yet to be systematically integrated into a unified theoretical framework. We attribute the former to the bargaining power mechanism underpinned by resource dependence theory and anchor the latter within the governance capability domain defined by dynamic capabilities theory, thereby consolidating fragmented insights into a theoretically consistent dual-channel explanatory framework. This dual-channel mechanism addresses the gap in existing resilience research, which predominantly focuses on macro levels while lacking micro-level transmission pathways. Therefore, this paper focus on bargaining power and governance capability.

First, according to the resource-dependence theory, bargaining power stems from the power asymmetry between trading partners [42]. Bargaining power, as the core of an enterprise’s vertical industrial chain relationship, directly influences the distribution of economic benefits in transactions [43]. However, the intensification of external environmental uncertainties increases the risk of industrial chain disruptions, which may weaken an enterprise’s bargaining power [38,44]. Enterprises mitigate industrial chain risks through supplier diversification, which reduces single-supplier dependence [39]. This strategy simultaneously enhances bargaining power along the value chain and strengthens price-setting capabilities in input markets. Specifically, digitalization enhances an enterprise’s bargaining power in two ways. On the one hand, it reduces information asymmetry and external transaction costs through information sharing [45]. On the other hand, it improves the collaborative efficiency between the enterprise and its suppliers [46]. By leveraging digital tools to optimize resource allocation and enhancing the enterprise’s coordination and management capabilities in complex industrial chains, it strengthens the enterprise’s bargaining power, thereby strengthening the industrial chain resilience. Therefore, we propose Hypothesis 1:

Hypothesis 1. 

Enterprise digitization can enhance the resilience of the industrial chain by improving bargaining power with upstream and downstream enterprises.

Second, according to dynamic capabilities theory [47], the essence of governance capacity lies in an enterprise’s ability to integrate, build, and reconfigure internal and external resources to adapt to dynamic environments. Through digitalization, enterprises enhance their dynamic capabilities, thereby directly improving the synergistic optimization of innovation efficiency and production efficiency within governance capacity. This strengthens the industrial chain resilience. At the innovation level, digitalization not only enables enterprises to assess technological feasibility and profit margins by leveraging market demand insights [48] and improves iterative efficiency in research and development to elevate output quality [41], but also facilitates cross-organizational knowledge sharing and collaborative innovation through digital platforms, accelerating technological diffusion and breakthrough achievements [49]. Enterprises’ innovation capabilities generate positive knowledge externalities that vertically integrate upstream suppliers and downstream customers, leading to coordinated productivity improvements throughout the industrial chain [50], which in turn reinforces systemic resilience against disruptions. At the managerial level, digitalization enhances information transparency and traceability, curbing information asymmetry and agency costs, while promoting the transition toward flatter, networked governance structures [51,52]. Concurrently, digitalization replaces traditional rigid production lines with flexible systems, enabling dynamic resource allocation and precise capacity matching, which reduces inventory costs and shortens production cycles [40,53]. These enhancements in managerial capabilities empower enterprises to respond swiftly to market demand fluctuations [54], thereby strengthening the industrial chain resilience. Therefore, we propose Hypothesis 2:

Hypothesis 2. 

Digitalization of enterprises can enhance the resilience of the industrial chain by improving innovation efficiency and production efficiency.

The theoretical framework is shown in Figure 1.

3. Data and Methods

3.1. Data Sources

The global financial crisis from 2008 to 2010 exerted a severe impact on the world economy, during which data may have been subject to short-term anomalous fluctuations, resulting in a lack of stability [55]. Beginning in 2011, the global economy gradually entered a recovery phase, with economic development stabilizing over time. This period provides more robust and reliable data for our analysis. Therefore, this paper selects the research window period from 2011 to 2023. The data mainly come from the CSMAR (China Stock Market & Accounting Research Database). The industrial chain resilience is mainly manifested in manufacturing enterprises, we only retain the data of manufacturing enterprises among the normally listed enterprises in the CSMAR database from 2011 to 2023, that is the listed enterprises with the industry code “C.” To improve the quality of the data, we exclude observation of ST, *ST, PT, and delisted companies. Meanwhile, we also exclude enterprise samples with incomplete enterprise information records and severely missing financial information.

Additionally, we utilized corporate annual report text data and China City Statistical Yearbooks. The corporate annual report text data was used to extract information on corporate digitization. We also extracted the annual regional GDP (gross domestic product) growth rates, the tertiary sector’s share of GDP, and real per capita GDP (2011 constant prices) from the city statistical yearbooks for the analysis of omitted variables. For the integration of multi-source data, we first merged the CSMAR data with the corporate annual report text data using unique stock codes. Next, we merged the data with the China City Statistical Yearbooks using the city codes where the enterprises are located. Finally, after data cleaning and integration, we retained a total of 28,939 unbalanced panel data points for enterprise-years from 2011 to 2023.

3.2. Variables

3.2.1. Industrial Chain Resilience

Industrial chain resilience refers to its ability to adapt to internal and external shocks [15,16]. Existing studies have not reached a fully consistent view on the definition and connotation of industrial chain resilience, and there are also significant differences in measurement methods [6,56,57,58]. Accordingly, drawing on relevant literature [20,59,60], we measure industrial chain resilience from resistance capability, operational capability, supply–demand matching capability, and renewal capability. The TOPSIS-entropy method combined the advantages of the entropy and the TOPSIS [61]. Specifically, the entropy-weighted TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method is a multi-criteria decision analysis approach that combines the entropy method and the TOPSIS method. Its fundamental principle involves determining the weights of each indicator through the entropy method [61], and then calculating the relative closeness of each alternative to the ideal solution and the negative ideal solution using the TOPSIS method [62], thereby ranking and evaluating the alternatives [63]. This method is widely applied to multi-indicator, multi-alternative decision-making problems, particularly in situations where conflicts exist between indicators or where there are inconsistencies in measurement units [20]. Therefore, we utilized the Entropy-Weighted TOPSIS method to assign weights and compute the industrial chain resilience by drawing on the indicator system (the basic principles and calculation steps of the Entropy Weighted TOPSIS method are shown in Appendix A.1).

For the indicator system, specifically, resistance capability emphasizes the ability to return to the original state when facing shocks [64]. Cash reserves are helpful in dealing with uncertainties [65,66]. We use the ratio of accounts receivable to main business revenue and the ratio of cash assets to total assets to measure the resistance capability of the industrial chain. Supply–demand matching capability emphasizes the balance between supply and demand. An improvement in logistics efficiency can increase the inventory turnover rate of enterprises, thereby effectively reducing their operating costs [67]. We use the ratio of the enterprise’s inventory value to total operating revenue and the inventory turnover rate to measure the supply–demand matching capability of the industrial chain. Operational capability represents an enterprise’s organizational and livelihood-earning capabilities [68], and it mainly enhances the enterprise’s management efficiency by accelerating capital flow and capital reserve [69]. We use the accounts receivable turnover rate and the accounts payable turnover rate to gauge the operational capacity of the industrial chain. Competitiveness highlights the capacity for renewal and upgrading. Innovation constitutes a crucial element of an enterprise’s core competitiveness [70], and investment in innovation aids in offering fundamental support for the research of key technologies [71]. We choose the proportion of research and development (R&D) staff to the total number of employees and the proportion of R&D expenditure to total operating income to measure the competitiveness of the industrial chain.

Table 1. Index system of industrial chain resilience.

Secondary Indicators		Third Indicators	Attribute	Weight
$ICR\_rc$	Resistance capability	The ratio of accounts receivable to main business revenue	−	0.024
		The ratio of cash to total assets	+	0.011
$ICR\_oc$	Operational capability	Accounts receivable turnover rate	+	0.025
		Accounts payable turnover ratio	−	0.121
$ICR\_sdmc$	Supply–demand matching capability	The ratio of inventory to main business revenue	−	0.119
		Inventory turnover rate	+	0.234
$ICR\_cc$	Competitive capability	The ratio of R&D personnel to total employees	+	0.164
		The ratio of R&D investment to main business revenue	+	0.302

Notes: (1) “−” represents a negative indicator, meaning that the larger the value, the lower the resilience of the industry chain. “+” represents a positive indicator, meaning that the larger the value, the higher the resilience of the industry chain. (2) The calculation process and results of weights are shown in Appendix A.1.

presents the definitions and attributes of the specific measurement indicators for industrial chain resilience.

3.2.2. Enterprise Digitalization Level

Drawing on the existing research [72], we measure enterprise-level digitalization by counting occurrences of 76 digital technology terms from artificial intelligence, big data, cloud computing, blockchain, and general digital technology applications (the specific terminology and vocabulary are as detailed in Appendix A.2). Subsequently, to address the issue of right-skewness, the natural logarithm transformation was applied to the data obtained from the frequency counts.

Figure 2 displays the relationship between enterprise digitalization and industrial chain resilience, with a fitted linear trend line. It could find that the slope of the linear fitting curve between them being 0.014, and as the level of enterprise digitalization advances, the industrial chain resilience exhibits a gradual increase. This presentation of factual data sets a realistic foundation for the empirical findings of this paper. To conduct a more in-depth exploration of the correlation between enterprise digitalization and industrial chain resilience, scatter plots were also constructed between enterprise digitalization and the secondary—dimension indicators of industrial chain resilience (as shown in Appendix A.3). The analysis shows that enterprise digitalization is positively correlated with resistance capability, supply–demand matching capability, and competitive capability. Conversely, a negative correlation is observed between enterprise digitalization and operational capability. These findings offer factual backing for our subsequent empirical examination of the relationship and the underlying mechanism between enterprise digitalization and industrial chain resilience.

3.2.3. Controls

Drawing on the research on industrial chains [6,71,73,74], these covariates include main business revenue ($revenue$), enterprise listing age ($age$), the proportion of tangible assets ($tangible$), Tobin’s Q ($tobin$), and the number of employees ($staff$). We conducted a 1% bilateral winsorization on all variables except age by year and industry. The definitions and descriptive statistics of the variables are shown in

Table 2. Definition and descriptive statistics of variables.

Variable	Definition	N	Mean	Std.
$ICR$	Industrial chain resilience	28,939	0.210	0.090
$digit$	Enterprise digitalization level	28,939	1.337	1.290
$revenue$	Logarithmic of main business revenue	28,939	21.989	1.191
$age$	Enterprise listing age	28,939	8.936	7.575
$tangible$	Proportion of tangible assets to total assets	28,939	0.353	0.151
$tobin$	The ratio of a company’s market value to its reset value	28,939	2.131	2.139
$staff$	Logarithmic number of employees	28,939	7.583	1.175

Notes: (1) To prevent the extreme-value effect, the $digit$, $revenue$, $age$, $tangible$, $tobin$, $staff$ are bilateral tail reduction at the 1% level by year and industry.

.

3.3. Empirical Model

We estimate the effect of enterprise digitalization on industrial chain resilience using OLS, following established empirical practice in industrial organization research [56,57,71]. The baseline specification takes the form as Equation (1).

$${ICR}_{ist}=\alpha +\beta {digit}_{ist}+\phi {control}_{ist}+\mu {industry}_s+\delta {year}_t+{\epsilon }_{ist}$$

(1)

In Equation (1), $i=\left(1, 2, \dots , 3908\right)$, $s=\left(1, 2, \dots , 30\right)$ and $t=\left(2011, 2012, \dots , 2023\right)$. ${ICR}_{ist}$ represents the industrial chain resilience for enterprise $i$ in industry $s$ in year $t$. ${digit}_{ist}$ denotes the digitalization level for enterprise $i$ in industry $s$ in year $t$. ${control}_{ist}$ is the set of covariates in this study, including $revenue$, $age$, $tangible$, $tobin$, and $staff$. Notably, disparities exist in the enterprise digitalization level among different industries and across various years (the distributions of the enterprise digitalization level by industry and year are presented in Appendix A.4). Consequently, it is imperative to account for industry-specific and year-specific biases in Equation (1). ${industry}_s$, ${year}_t$, and ${\epsilon }_{ist}$ represents industry fixed effects, year fixed effects, and error term, respectively.

4. Results

4.1. Baseline Estimation Results

We employ Equation (1) to examine the relationship between enterprise digitalization and industrial chain resilience, and the results are shown in columns 1–2 of

Table 3. Impact of enterprise digitization on industrial chain resilience.

	(1)	(2)	(3)	(4)	(5)	(6)
	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}\_\mathit{r}\mathit{c}$	$\mathit{I}\mathit{C}\mathit{R}\_\mathit{o}\mathit{c}$	$\mathit{I}\mathit{C}\mathit{R}\_\mathit{s}\mathit{d}\mathit{m}\mathit{c}$	$\mathit{I}\mathit{C}\mathit{R}\_\mathit{c}\mathit{c}$
$digit$	0.007 ***	0.008 ***	−0.000	−0.000	0.001 ***	0.010 ***
	(0.001)	(0.001)	(0.000)	(0.000)	(0.000)	(0.001)
$revenue$		0.019 ***	−0.003 ***	0.004 ***	0.008 ***	0.017 ***
		(0.002)	(0.000)	(0.001)	(0.001)	(0.002)
$age$		−0.002 ***	−0.000 ***	0.000 ***	−0.000 ***	−0.001 ***
		(0.000)	(0.000)	(0.000)	(0.000)	(0.000)
$tangible$		0.012 *	−0.064 ***	0.022 ***	0.082 ***	−0.045 ***
		(0.007)	(0.002)	(0.004)	(0.003)	(0.006)
$tobin$		0.002 ***	0.000 ***	0.001 ***	0.000	0.004 ***
		(0.001)	(0.000)	(0.000)	(0.000)	(0.001)
$staff$		−0.023 ***	0.002 ***	0.001 *	−0.012 ***	−0.026 ***
		(0.002)	(0.000)	(0.001)	(0.001)	(0.002)
Constant	0.201 ***	−0.033	0.112 ***	−0.068 ***	−0.069 ***	−0.056 **
	(0.001)	(0.028)	(0.008)	(0.015)	(0.012)	(0.028)
Industry fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
N	28,938	28,938	28,931	28,826	28,904	18,616
R2	0.330	0.373	0.216	0.334	0.300	0.419

Notes: (1) The robust standard errors for clustering at the enterprise level are shown in parentheses. (2) *, **, and *** represent significance at the 10%, 5%, and 1% statistical levels, respectively.

. The coefficient on $digit$ remains positive and statistically significant both with and without the inclusion of control variables. For instance, as demonstrated in column 2, the coefficient for enterprise digitalization is estimated at 0.008, achieving significance at the 1% level. The estimates indicate that a 1% increase in enterprise digitalization leads to a 0.008-unit enhancement in industrial chain resilience. This implies that for every 1% rise in digitalization levels, industrial chain resilience increases by 3.8% relative to the average level. This economically significant effect suggests digitalization plays a crucial role in enhancing industrial chain resilience.

Building upon this foundation, we further investigated the influence of enterprise digitalization on the secondary indicators of industrial chain resilience through Equation (1). The empirical findings, as presented in columns 3–6 of Table 3, indicate that enterprise digitalization does not exhibit a statistically significant effect on resistance capability and operational capability. However, it demonstrates a markedly positive impact on both supply–demand matching capability and competitive capability. These results suggest that enterprise digitalization primarily contributes to the enhancement of industrial chain resilience by improving the supply–demand matching capability and competitive capability strength within the industrial chain.

In addition, to avoid the spurious estimation results that may arise from a decrease in main business income and total assets, leading to an increase in industrial chain resilience, we also estimated the impact of enterprise digitization on main business income and total assets (the estimation results are presented in Appendix A.5). The results show that digitization has a positive effect on both main business income and total assets, indicating that the aforementioned potential spurious estimation results do not exist.

In addition, we further examined the influence of enterprise digitalization on specific indicators of industrial chain resilience, with the empirical results depicted in Figure 3. It reveals that the enterprise digitalization does not exert a statistically significant effect on the ratio of accounts receivable to total operating revenue, the ratio of cash to total assets, the accounts receivable turnover rate, or the inventory turnover rate. However, enterprise digitalization significantly reduces the accounts payable turnover rate and the ratio of inventory value to total operating revenue. Conversely, it significantly enhances the proportion of R&D personnel relative to the total workforce and the ratio of R&D investment to total operating revenue. Given that both the accounts payable turnover rate and the ratio of inventory value to total operating revenue are inverse indicators, enterprise digitalization positively influences industrial chain resilience primarily by decreasing these two metrics while simultaneously increasing the proportion of R&D personnel relative to the total workforce and the ratio of R&D investment to total operating revenue.

4.2. Robustness Test

4.2.1. Replace $ICR$

The industrial chain risk ($ICS$) captures the extent of misalignment between production and demand within the industrial chain, as orchestrated by the core enterprise. Building on existing research, we operationalize $ICS$ as the ratio of the quarterly standard deviation of production volume to the quarterly standard deviation of demand volume [73]. Specifically, we use the sum of quarterly cost of sales and quarterly net inventory as the quarterly production volume indicator, and the sum of quarterly main business income and other business income as the quarterly demand volume. Before calculating the standard deviation, we perform a logarithmic transformation on the quarterly production volume and quarterly demand volume. We substitute the industrial chain resilience ($ICR$) with the industrial chain risk ($ICS$). The estimated results are shown in the column 1 of

Table 4. The results of robustness Test.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
	$\mathit{I}\mathit{C}\mathit{S}$	$\mathit{I}\mathit{C}\mathit{R}\_\mathit{f}\mathit{a}\mathit{c}\mathit{t}\mathit{o}\mathit{r}$	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$
$digit$	−0.014 **	0.049 ***				0.003 ***	0.010 ***
	(0.006)	(0.005)				(0.001)	(0.001)
$digit\_99$			0.011 ***
			(0.001)
$digit\_139$				0.010 ***
				(0.001)
$Broadband\_China$					0.013 ***
					(0.002)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes	Yes	No	Yes
Enterprise fixed effect	No	No	No	No	No	Yes	No
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes	Yes
N	26,708	28,938	28,937	28,938	25,272	28,645	26,037
R2	0.006	0.600	0.329	0.332	0.330	0.739	0.330

Notes: (1) The robust standard errors for clustering at the enterprise level are shown in parentheses. (2) ** and *** represent significance at the 5%, and 1% statistical levels, respectively. (3) Controls include $revenue$, $age$, $tangible$, $tobin$, and $staff$.

, indicating that enterprise digitization significantly reduces industry chain risks. Industrial chain risk ($ICS$) refers to the potential negative impacts of internal and external shocks on the industrial chain system [73], while industrial chain resilience denotes the system’s ability to recover from such shocks [15,16,64]. Thus, according to their definitions, a reduction in industrial chain risk can enhance industrial chain resilience.

4.2.2. Replace the Weighting Method

Factor analysis is widely adopted weighting approach, and it also was employed to reallocate weights within the industrial chain resilience index system [75]. Through this method, a revised industrial chain resilience ($ICR\_factor$) was computed via weighted aggregation. Then we re-estimated Equation (1), and the results are shown in column 2 of Table 4. The enterprises’ digitization still significantly enhances the industrial chain resilience.

4.2.3. Replace the Digital Word Frequency Library

In existing studies, prevalent approaches for quantifying digitalization also involve measuring the occurrence rates of 99 digital-relevant terms from digital technology implementation, internet-based business frameworks, intelligent production systems, and contemporary information architectures [74], while others involve enumerating 139 digital-associated terms categorized under technical taxonomies, organizational capacity enhancement, and digital operational practices [76]. Consequently, logarithmic transformations were applied to the word frequency metrics of the two enterprise digitalization indicators, respectively designated as $digit\_99$ and $digit\_139$. Then we re-estimated Equation (1), and the results are shown in column 3–4 of Table 4. The enterprises’ digitization still has a significant positive impact on the industrial chain resilience.

4.2.4. Replace Enterprise Digitalization

Digital infrastructure constitutes the cornerstone for the enterprises digitization, and then has the potential to reconfigure existing industrial chains, enhancing their safety, stability, and operational efficiency [77]. Furthermore, some studies have used the “Broadband China” policy as a digital shock to estimate its impact on the industrial chain [78]. Similarly, we also use the “Broadband China” policy as a digital shock ($Broadband\_China$) and employ the difference-in-differences (DID) method to estimate the impact of digitization on the industrial chain resilience¹. The results are shown in column 5 of Table 4, indicating that the “Broadband China” policy has markedly bolstered the industrial chain resilience, thereby corroborating the robustness of the baseline estimation.

4.2.5. Using Firm Fixed Effects

To control for time-invariant, unobserved firm characteristics that may be related to both digitalization and industrial chain resilience, we further employ firm fixed effects. We re-estimate the model using two-way fixed effects for both firm and year. The estimation results are shown in Column 6 of Table 4. The impact of firm digitalization on industrial chain resilience remains significantly positive, consistent with the benchmark estimation results.

4.2.6. Propensity Score Matching (PSM)

To mitigate potential sample selection bias arising from the incidental nature of firms’ digital transformation, we further conduct PSM by matching non-digitalized firms with digitalized firms using a 1-to-1 matching method within a caliper of 0.05². The model is re-estimated using the matched sample. The results, presented in Column 7 of Table 4, indicate that after addressing sample selection bias, the positive impact of firm digitalization on industrial chain resilience remains statistically significant.

4.2.7. Year-by-Year Estimation

To more rigorously test the robustness of the benchmark estimation results and to exclude the potential influences of economic recovery in the early stage of the sample period and public health incident shocks in the later stage, we conduct year-by-year estimations within the research window. The estimation results, as illustrated in Figure 4, show that the effect of firm digitalization on industrial chain resilience is significantly positive in each year throughout the research window. This result helps rule out potential interference from other factors within the sample period.

4.3. Endogenous Analysis

4.3.1. Omitted Variable

The enterprise digitalization and the robustness of industrial chains are both contingent on macroeconomic conditions [8]. Our initial specification omitted city and province level controls, potentially introducing omitted variable bias. To mitigate this endogeneity concern, we incorporate annual regional GDP growth rates, the tertiary sector’s share of GDP, and real per capita GDP (2011 constant prices). These variables serve as comprehensive proxies for regional economic development. We sequentially introduce city and province level macroeconomic controls into Equation (1), and the results are shown in columns 1–2 of

Table 5. The results of Endogenous analysis.

	(1)	(2)	(3)	(4)	(5)	(6)
	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$
$digit$	0.008 ***	0.008 ***		0.038 ***		0.068 ***
	(0.001)	(0.001)		(0.007)		(0.009)
$IV1$			0.066 ***
			(0.006)
$IV2$					0.004 ***
					(0.000)
Kleibergen–Paap rk LM statistic			115.536		107.628
Kleibergen–Paap Wald rk F statistic			116.330		107.606
Controls	Yes	Yes	Yes	Yes	Yes	Yes
City Controls	Yes	Yes	No	No	No	No
Province Controls	No	Yes	No	No	No	No
Industry fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes	Yes	Yes
N	25,268	25,268	23,784	23,784	23,784	23,784
R2	0.386	0.386		0.279		−0.060

Notes: (1) The robust standard errors for clustering at the enterprise level are shown in parentheses. (2) *** represent significance at the 1% statistical levels, respectively. (3) The Kleiberen Paap rk LM statistics for the unidentifiable tests of IV1 and IV2 are 115.536 and 107.628, respectively, while the Kleiberen Paap Wald rk F statistics for the weakly recognizable tests of IV1 and IV2 are 116.330 and 107.606, respectively, both significantly above the critical value of 16.38. (4) Controls include $revenue$, $age$, $tangible$, $tobin$, and $staff$. (5) The R² in columns (4) and (6) is the R² after centralization.

. After considering these additional control measures, the positive and statistically significant impact of enterprise digitization on industry chain resilience still exists. It is worth noting that the estimated coefficients are still similar in quantity to our baseline specifications, indicating that the omitted variable bias in our baseline results is very small.

4.3.2. IV Estimates

We use the frequency of words in the annual report of an enterprise to measure the endogeneity of measurement errors that may arise from enterprise digitalization, leading to inaccurate results. We adopted the instrumental variable method to obtain more accurate estimation results. The number of post offices per million people and the number of fixed telephones per 10 thousand people in 1984 are often used as instrumental variables of digitalization [79,80,81]. These instrumental variables satisfy the relevance condition, as regions with greater telecommunications infrastructure development historically exhibit higher probabilities of subsequent enterprise digitalization. However, early communication infrastructure reflects historical state planning or geographical constraints. Although it may have influenced regional development trajectories in the past, it cannot directly affect modern industrial chains. Moreover, research emphasizes that historical instrumental variables can capture long-term institutional or geographical differences without being correlated with contemporary short-term shocks [82]. Therefore, the number of post offices per million people and the number of fixed telephones per ten thousand people in 1984 serve as suitable instrumental variables. Therefore, we adopt the method of dealing with the time trend of instrumental variables and constructed $IV1$ by interacting the number of post offices per million people in 1984 with the year variable, and $IV2$ by interacting the number of fixed telephones per ten thousand people in 1984 with the year variable. $IV1$ and $IV2$ are the instrumental variables for enterprise digitalization. Then, we employ a two-stage least squares (2SLS) method to estimate the effect of enterprise digitalization on industrial chain resilience.

The results are shown in columns 3–6 of Table 5. Columns 3 and 5 report the first-stage estimates for $IV1$ and $IV2$, respectively. Both instrumental variables demonstrate statistically significant positive effects on enterprise digitalization, satisfying the relevance condition for valid instruments. Columns 2 and 4 report the second-stage results, and reveal substantial and statistically significant effects of enterprise digitalization ($digit$) on industrial chain resilience, with estimated coefficients of 0.038 for IV1 and 0.068 for IV2. These effect sizes notably exceed the corresponding OLS estimate from column (2) of Table 3, suggesting that conventional regression approaches may substantially underestimate digitalization’s true impact. This pattern of results serves two important purposes. First, it establishes that our baseline findings are robust to endogeneity concerns when using plausibly exogenous variation. Second, the fact that the 2SLS coefficients are consistently larger than the OLS estimates suggests that standard estimation methods may attenuate the true treatment effects because of measurement error.

4.4. Heterogeneity Analysis

The impact of enterprise digitization on industrial chain resilience may exhibit heterogeneity among enterprise attributes. Firstly, enterprises of varying scales exhibit marked differences in managerial costs, production structures, and market demand characteristics. This pattern implies that enterprise size likely moderates the digitalization–resilience relationship. We proxy enterprise size using total assets and introduce an interaction term between enterprise size and digitalization into Equation (1). The results in column 1 of

Table 6. The results of heterogeneity analysis.

	(1)	(2)	(3)	(4)
	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$	$\mathit{I}\mathit{C}\mathit{R}$
$digit$	0.056 ***	0.034 ***	0.001	0.009 ***
	(0.012)	(0.010)	(0.003)	(0.001)
$size$	−0.016 ***
	(0.002)
$digit\times size$	−0.002 ***
	(0.001)
$capital\_intensity$		−0.021 ***
		(0.002)
$digit\times capital\_intensity$		−0.002 ***
		(0.001)
$ESG$			0.006 ***
			(0.001)
$digit\times ESG$			0.001 *
			(0.001)
$owner$				0.018 ***
				(0.004)
$digit\times owner$				−0.001
				(0.002)
Controls	Yes	Yes	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes
N	28,932	28,935	27,889	25,040
R2	0.387	0.395	0.380	0.369

Notes: (1) The robust standard errors for clustering at the enterprise level are shown in parentheses. (2) * and *** represent significance at the 10% and 1% statistical levels, respectively. (3) Controls include $revenue$, $age$, $tangible$, $tobin$, and $staff$.

reveal a statistically significant negative coefficient for the interaction term $digit\times size$, suggesting that digitalization has brought greater industrial chain resilience to small enterprises compared to large enterprises. This may be due to the more flexible operations of small enterprises [83].

Second, relative to capital-intensive enterprises, enterprise digitalization appears to exert a more pronounced effect on strengthening risk resilience in labor-intensive enterprises [84]. To examine this heterogeneity, we construct a capital intensity index ($capital\_intensity$) using the ratio of net fixed asset value to total employment. Then we introduce an interaction term between capital intensity and digitalization into Equation (1). The results are shown in column 2 of Table 6, the interaction term $digit\times capital\_intensity$ exhibits a significant negative coefficient, suggesting that digitalization has a greater impact on the industry chain resilience of enterprises with lower capital intensity compared to those with higher capital intensity.

Third, a robust ESG (environmental, social, and governance) rating serves as both an indicator of an enterprises’ commitment to social responsibility and a proxy for substantial capital market endorsement [85]. Enterprises with elevated ESG ratings exhibit greater capacity to drive digital innovation and expedite their digital transformation processes [86]. To investigate this relationship, we introduce an enterprise’s annual ESG score as a moderating variable and include its interaction with enterprise digitalization into Equation (1). The results are shown in column 3 of Table 6, the interaction term $digit\times ESG$ exhibits a significant positive coefficient, indicating that digitalization generates more substantial industrial chain resilience improvements for enterprises with higher ESG ratings. The reason may be that enterprises with higher ESG ratings demonstrate more efficient resource allocation, stricter internal controls [87], and the ability to access capital at lower costs and attract sustainable investment, providing stable funding support for digital infrastructure [88]. This results in a stronger impact of digitalization on the industrial chain resilience.

Finally, there are significant differences in production efficiency and business objectives between state-owned and private enterprises in China [89]. The impact of digitization on the industrial chain resilience may vary depending on ownership structure. Following the existing research [90], we classify enterprises into state-owned enterprises and private enterprises based on their registration types. We construct a binary variable ($owner$), assigning a value of 1 to state-owned enterprises and 0 to private enterprises, and introduce its interaction with enterprise digitalization into Equation (1). The results are shown in column 4 of Table 6, the interaction term $digit\times owner$ is insignificant, indicating that the resilience-enhancing effect of digitalization does not differ between state-owned enterprises and private enterprises.

4.5. Mechanism Analysis

Based on the theoretical analysis, we focus on testing the bargaining power mechanism and the governance capability mechanism. Focusing first on bargaining power, enterprise digitalization enables industrial chain diversification, reducing single-supplier dependence and strengthening negotiation leverage [39]. Therefore, we use the upstream supplier concentration dispersion³ ($up\_SCD$) and downstream customer concentration dispersion ($down\_SCD$) of enterprises as proxy indicators for the enterprise bargaining power mechanism. We employ the OLS model to confirm this mechanism, and the estimation results are shown in columns 1–2 of

Table 7. The results of mechanism analysis.

	(1)	(2)	(3)	(4)
	Bargaining Power		Governance Capability
	$\mathit{u}\mathit{p}\_\mathit{S}\mathit{C}\mathit{D}$	$\mathit{d}\mathit{o}\mathit{w}\mathit{n}\_\mathit{S}\mathit{C}\mathit{D}$	$\mathit{i}\mathit{n}\mathit{n}\mathit{o}\mathit{v}\mathit{a}\mathit{t}\mathit{e}$	$\mathit{T}\mathit{F}\mathit{P}\_\mathit{O}\mathit{P}$
$digit$	1.157 ***	1.830 ***	0.006 ***	0.028 ***
	(0.187)	(0.233)	(0.002)	(0.006)
Controls	Yes	Yes	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes
N	26,105	28,198	28,938	26,399
R2	0.286	0.241	0.054	0.628

Notes: (1) The robust standard errors for clustering at the enterprise level are shown in parentheses. (2) *** represent significance at the 1% statistical levels, respectively. (3) Controls include $revenue$, $age$, $tangible$, $tobin$, and $staff$.

. The positive and statistically significant coefficients on digitalization demonstrate that digital adoption reduces industrial chain concentration by 1.16% upstream and 1.83% downstream. The digitalization enhances the bargaining power of enterprises over upstream and downstream partners as theoretically predicted. These estimates support the bargaining power channel, confirmed our Hypothesis 1.

Existing research establishes that both innovation efficiency and production efficiency contribute to industrial chain resilience [13,91]. We use the ratio of the number of an enterprise’s invention patents to the total number of patents as a proxy variable for enterprise innovation efficiency ($innovate$), and the total factor productivity of the enterprise calculated by the OP method [92] as a proxy variable for production efficiency ($TFP\_OP$). We employ the OLS model to confirm this mechanism, and the results are shown in columns 3–4 of Table 7. The results demonstrate statistically significant positive coefficients of enterprise digitalization on both innovation efficiency and production efficiency. The digitalization enhances innovation output by improving patent quality composition, boosts productive efficiency, and then consequently strengthens corporate governance capacity, as theoretically predicted. The results presented in Table 7 collectively validate our theoretical framework, demonstrating that digitalization affects industrial chain resilience through both enhanced bargaining power and improved governance capabilities, confirmed our Hypothesis 2.

5. Spillover Effects of the Industrial Chain

Corporate digitalization serves multiple critical functions in enhancing the supply-chain ecosystem. It empowers enterprises to construct industrial chain information platforms, which are instrumental in preventing suboptimal production planning and bolstering the stability of the industrial chain [5]. In addition, digitalization enables enterprises to swiftly gather customer requirements and seamlessly integrate these consumer demands into key business processes, including design, manufacturing, and marketing services. This integration not only enhances customer satisfaction but also strengthens industrial chain resilience [17]. Given these advantages, corporate digitalization not only directly bolsters an enterprise’s industrial chain resilience but also generates spillover benefits across upstream suppliers and downstream customers within the industrial chain network [93].

To empirically evaluate the spillover effects of corporate digitalization on industrial chain resilience, we employed an annual matching based on enterprises’ procurement and sales data. Finally, we have obtained the mixed panel data for 2584 core enterprises and their upstream suppliers, as well as 3978 core enterprises and their downstream customers (The matching process and descriptive statistics of the data are provided in Appendix A.10). We employ the OLS model to analyze the spillover effects of core enterprise digitalization across upstream suppliers and downstream customers. The results are shown in

Table 8. Spillover effects of enterprise digitization on upstream and downstream enterprises.

	(1)	(2)	(4)	(5)
	$\mathit{d}\mathit{i}\mathit{g}\mathit{i}\mathit{t}\_\mathit{u}\mathit{p}$	$\mathit{I}\mathit{C}\mathit{R}\_\mathit{u}\mathit{p}$	$\mathit{d}\mathit{i}\mathit{g}\mathit{i}\mathit{t}\_\mathit{d}\mathit{o}\mathit{w}\mathit{n}$	$\mathit{I}\mathit{C}\mathit{R}\_\mathit{d}\mathit{o}\mathit{w}\mathit{n}$
$digit$	0.322 ***	0.002	0.396 ***	0.008 ***
	(0.038)	(0.003)	(0.039)	(0.002)
Controls	Yes	Yes	Yes	Yes
Up enterprises controls	Yes	Yes	No	No
Down enterprises controls	No	No	Yes	Yes
Industry fixed effect	Yes	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes	Yes
N	2584	2584	3978	3978
R2	0.363	0.239	0.504	0.255

Notes: (1) The robust standard errors for clustering at the enterprise level are shown in parentheses. (2) *** represent significance at the 1% statistical levels, respectively. (3) Controls include $revenue$, $age$, $tangible$, $tobin$, and $staff$. Up enterprises controls include $revenue$, $age$, $tangible$, $tobin$, and $staff$ of upstream suppliers. Down enterprises controls include $revenue$ of upstream suppliers. Down enterprises controls include $revenue$, $age$, $tangible$, $tobin$, and $staff$ of downstream customers.

.

Columns 1–2 of Table 8 show the impacts of the core enterprise digitalization ($digit$) on the upstream supplier digitalization ($digit\_up$) and the industrial chain resilience of upstream suppliers ($ICR\_up$). The digitalization of core enterprises significantly enhances the digitalization level of upstream suppliers. However, its effect on improving the industrial chain resilience of upstream suppliers is not significant. Columns 3–4 of Table 8 show the impacts of the digitalization of core enterprises ($digit$) on the downstream customers digitalization ($digit\_down$) and the industrial chain resilience of downstream customers ($ICR\_down$). The digitalization of core enterprises significantly boosts both the digitalization level and the industrial chain resilience of downstream customers. The findings suggest that corporate digitalization can drive both upstream suppliers and downstream customers to embark on digitalization. However, the spillover effect of corporate digitalization on industrial chain resilience primarily manifests in the backward extension of the supply chain, with no discernible spillover effect in the forward extension. This heterogeneity may be related to supply chain power, where downstream enterprises play a dominant role in the industry chain compared to upstream suppliers [21].

6. Discussion and Conclusions

This paper examines how enterprise digitalization enhances industrial chain resilience in the manufacturing sector. Using panel data from 2011 to 2023, we employ an OLS model to estimate this relationship. Our findings indicate that enterprise digitalization significantly improves industrial chain resilience, primarily by strengthening supply–demand matching capability and competitive capability. These results remain robust after addressing endogeneity concerns and conducting extensive sensitivity tests. Heterogeneity analysis reveals that the effect varies by enterprise size, capital intensity, and ESG performance, but shows no significant difference between state-owned and private enterprises. Mechanism testing further indicates that the improvement of bargaining power and governance capability is the main channel for the impact of enterprises digitization on industrial chain resilience. Specifically, on the one hand, enterprise digitization enhances bargaining power by promoting the diversification of supplier and customer networks. On the other hand, it strengthens governance capabilities by promoting innovation and improving production efficiency. Additionally, we identify significant backward spillover effects to upstream and downstream partners.

In contrast to existing research, which predominantly examines the relationship between digitalization and industrial chains from global and urban perspectives [15,21]. We advance the discussion in three key aspects. First, we extend the analysis of industrial chain resilience from a macro perspective to the micro-level foundations of enterprises. Second, we construct a dual-channel framework based on resource dependence theory and dynamic capabilities theory, revealing both the internal and external pathways through which digitalization enhances resilience. Finally, we introduce a directional spillover perspective that clarifies the asymmetric transmission mechanisms of digitalization along the industrial chain.

With our research findings, the impact of corporate digitalization on industrial chain resilience varies by enterprise type, with the strongest effects observed in small enterprises, labor-intensive enterprises, and those with higher ESG scores. However, we find no statistically significant difference between state-owned enterprises and private enterprises. This heterogeneity is essentially the result of the reconfiguration of digital capabilities shaped by the enterprise scale threshold, factor substitution elasticity, and institutional compliance threshold. Firstly, the advantages of small enterprises in production, operation, and management efficiency make digitalization have a greater impact on the industrial chain resilience [83]. Secondly, labor-intensive enterprises benefit from the super-linear substitution elasticity of digitalization for labor, transforming their traditional reliance on human labor into an advantage in digital process control [94]. Thirdly, high-ESG enterprises can obtain a premium for green supply chain certification through tools such as blockchain traceability, converting ESG data assets into institutional legitimacy capital to avoid trade barriers and financing constraints [85]. Enterprises with higher ESG ratings also exhibit more efficient resource allocation, stricter internal controls [87], and the ability to obtain capital at lower costs and attract sustainable investment [88]. This results in a stronger impact of digitalization on the industrial chain resilience. Fourth, the null result between state-owned and private enterprises reflects institutional convergence in digital economy regulation and policy homogenization across ownership types. Specifically, cross-cutting industrial chain security policies and standardized digital governance frameworks have neutralized traditional resource disparities based on property rights [95].

The spillover effects of enterprise digitalization primarily enhance industrial chain resilience through backward propagation along the value chain. While a core function of digitalization is to reduce information asymmetries [3], we observe directional heterogeneity in the information integration capabilities across different industrial chain resilience. Enterprises typically wield greater bargaining power over upstream suppliers [21]. digitalization enables real-time monitoring that compels process standardization among suppliers, creating a reverse coercive mechanism in industrial chain governance. However, downstream suppliers face greater market fragmentation and demand heterogeneity, making them less responsive to standardized digital governance. This market structure limits the forward propagation of resilience benefits through the industrial chain [96].

To effectively enhance industrial chain resilience, the development of digitalization requires coordinated efforts at both the enterprise and policy levels. At the enterprise level, companies should formulate a digitalization strategy centered on strengthening bargaining power and governance capabilities. Specific steps include: (1) Conducting a digital maturity assessment based on their own scale and industry characteristics to clarify the transformation pathway. (2) Setting quantifiable targets, such as reducing supplier concentration through digital procurement platforms, improving inventory turnover with IoT technologies, and enhancing patent quality by promoting digital R&D collaboration. (3) Building a self-centered digital ecosystem, empowering downstream partners through standardized interfaces and data sharing to amplify the positive spillover effects on resilience. (4) Establishing a composite indicator system that covers digitalization level, resilience performance, and mechanism effectiveness to enable real-time risk perception and dynamic optimization.

At the government level, policies should shift from general encouragement to targeted empowerment. (1) Implementing differentiated incentives, such as providing tool subsidies and training for SMEs, supporting labor-intensive enterprises in integrating “automation + data cloud” upgrades, and guiding high-ESG enterprises to develop green digital supply chains while connecting them with green finance resources. (2) Launching a “Digital Enablement Program for Core Enterprises,” offering tax benefits or project preferences to core companies that actively provide technology, data, or standards to their upstream and downstream partners, thereby encouraging them to drive industrial chain collaboration. (3) Establishing regional digital transformation promotion centers to provide neutral services in capability diagnosis, solution matching, and financing coordination, systematically enhancing enterprises’ digital absorption capacity. Through coordinated efforts between enterprises and the government, digitalization can be systematically embedded in the construction of industrial chain resilience, laying a solid foundation for building a safe, efficient, and collaborative modern industrial system.

Although we have systematically analyzed the relationship between enterprise digitalization and industrial chain resilience, its mechanisms, and externalities, our research still has certain limitations. For instance, despite the resumption of economic activities, the impact of COVID-19 cannot be entirely eliminated. Additionally, in a volatile trade environment, the conservative strategies adopted by some enterprises may lead to an overestimation of the effect of enterprise digitalization on industrial chain resilience in our findings.