A Study on the Impact of Environmental Penalties on Corporate Supply Chain Resilience

Abstract

Against the backdrop of increasingly stringent environmental regulation and increasing uncertainty in supply chain operations, this study examines how environmental penalties affect corporate supply chain resilience. Using Chinese A-share listed firms from 2009 to 2024, this paper constructs a firm-level panel dataset and employs a two-way fixed-effects model to estimate the relationship between environmental penalty intensity and supply chain resilience. Environmental penalty intensity is measured by the annual penalty amount imposed on each firm, while supply chain resilience is captured through an entropy-weighted index reflecting both resistance and recovery capacities. To alleviate endogeneity concerns, this study further uses an instrumental-variable approach based on the interaction between a firm’s one-year lagged penalty amount and city-level thermal inversion days. The results show that environmental penalties reduce corporate supply chain resilience. This negative effect is heterogeneous across firm characteristics and is partially mediated by reduced operational efficiency and crowded-out R&D investment. This conclusion remains robust after replacing the dependent variable, changing the clustering level of standard errors, and excluding observations from the COVID-19 pandemic period. Mechanism tests suggest that environmental penalties weaken supply chain resilience partly by reducing operational efficiency and crowding out R&D investment. Heterogeneity analysis indicates that the negative effect is more pronounced among young firms, non-high-tech firms, and firms located in regions with lower environmental regulation intensity. This study contributes to the literature by distinguishing environmental penalties from broader environmental regulation and by examining their implications for supply chain resilience. The findings also suggest that environmental enforcement should maintain deterrence while improving transparency, predictability, and targeted compliance guidance.

1. Introduction

In recent years, against the backdrop of continual adjustments to the global industrial chain and ongoing progress in green transformation, uncertainty in corporate operating environments has increased, and shocks to supply chain systems have become increasingly diverse. Under such external shocks, maintaining stable supply chain operations and achieving rapid recovery have gradually emerged as key concerns in corporate risk management and operations management. Supply chain resilience is commonly viewed as a critical indicator of a firm’s ability to withstand disruptions and recover from shocks (Ponomarov & Holcomb, 2009; Pettit et al., 2010) [1,2]. In the context of macroeconomic fluctuations and evolving institutional environments, strengthening corporate supply chain resilience is an essential condition for ensuring business continuity.

Meanwhile, as China’s ecological civilization construction continues to advance, the intensity of environmental supervision has steadily increased. Environmental regulation has gradually become an important institutional factor shaping firms’ business decisions. Existing research suggests that environmental regulation can promote corporate technological innovation and green transformation through a “reverse-forcing mechanism” (Ambec et al., 2013) [3]. In the Chinese context, a large body of empirical evidence shows that environmental regulation helps foster firms’ green technological innovation and improve their environmental governance behavior (Cui et al., 2022; Liu et al., 2024) [4,5]. In addition, some studies indicate that environmental regulation may also affect corporate performance, investment decisions, and market behavior by altering cost structures and resource allocation patterns (Lanoie et al., 2011) [6]. Overall, prior research has mainly examined the effects of environmental regulation on corporate green innovation, environmental governance, and operating performance.

It is important to distinguish environmental penalties from other dimensions of environmental regulation that are often discussed together in the literature. Environmental regulation may refer to formal environmental policies, such as low-carbon city pilots or carbon trading pilots; regional regulatory intensity, such as emission standards or environmental enforcement resources; disclosure of environmental violations, such as environmental blacklists or public penalty announcements; and actual monetary penalties imposed on firms for specific violations. These dimensions differ in their policy level, transmission mechanism, and empirical interpretation. Formal environmental policies usually operate at the macro or regional level; regulatory intensity captures local enforcement willingness and resources; violation disclosure mainly works through reputational channels; and the actual penalty amount directly reflects the monetary cost borne by the penalized firm. Therefore, this study treats the environmental penalty amount as a specific enforcement shock with observable intensity rather than as a general proxy for environmental regulation. This distinction helps clarify the theoretical boundary of the study and avoids conflating environmental penalties with broader environmental regulatory measures.

In supply chain management research, scholars have primarily examined the formation mechanisms of supply chain resilience from perspectives such as digital transformation, corporate governance, and the institutional environment. For instance, digital technologies can enhance supply chain stability by improving information sharing and resource allocation efficiency (Ivanov & Dolgui, 2021) [7]. Strong corporate governance and ESG performance can strengthen partner trust and thereby improve the supply chain’s ability to withstand shocks (Gölgeci & Kuivalainen, 2020) [8]. However, no consensus has been reached regarding the relationship between environmental regulation shocks and supply chain resilience. Two competing views are commonly discussed. The first is the “reverse-forcing enhancement” view, which argues that environmental penalties exert compliance deterrence and signaling effects. Through these channels, firms are pressured to strengthen environmental risk management, optimize supply chain cooperation and resource allocation, and thereby improve supply chain resilience and sustainability. The second is the “shock-hindering” view, which contends that environmental penalties impose penalty costs. These burdens crowd out R&D resources, reduce operational efficiency, and ultimately weaken supply chain stability and resilience. Overall, prior research has mainly focused on firms’ internal capabilities or the optimization of the institutional environment. However, empirical studies that directly test environmental penalties as a specific regulatory instrument remain limited. Moreover, existing studies have not systematically clarified the net effects and transmission mechanisms through which environmental penalties influence supply chain resilience. This gap provides an important entry point for the present study.

Compared with macro-level environmental policies, environmental penalties typically affect corporate business activities more directly through specific penalty events. They not only increase firms’ compliance costs but may also influence corporate reputation and the financing environment through information transmission mechanisms. When a firm receives a penalty, financial institutions and supply chain partners may reassess the firm’s risk, which in turn affects its resource allocation capability. As a result, the firm’s production and operational efficiency may decline, and its capacity for long-term R&D investment may be crowded out, thereby affecting the stability of the supply chain system. Therefore, systematically examining the effects of environmental penalties on corporate supply chain resilience at the micro level and further investigating the underlying mechanisms are of considerable significance for deepening research on the economic consequences of environmental regulation.

Based on the above discussion, this study uses Chinese A-share listed firms from 2009 to 2024 as the research sample. It empirically examines the effect of environmental penalties on corporate supply chain resilience and further tests the mediating roles of operational efficiency and R&D investment. The objective is to identify, at the microlevel, the internal mechanisms through which environmental regulation affects supply chain stability.

Building on existing research, this study may make the following marginal contributions.

First, it investigates the economic consequences of environmental penalties at the micro-enterprise level. Prior research mainly examines how environmental regulation affects corporate green innovation, environmental governance behavior, and operating performance. However, evidence on environmental penalties in the context of supply chain operations remains relatively limited. This study therefore extends the research perspective from corporate environmental governance behavior to supply chain management by examining how environmental penalties influence corporate supply chain resilience. In doing so, it enriches the literature on the economic consequences of environmental regulation.

Second, this paper develops and tests an analytical framework. The framework is “environmental penalties → operational efficiency, R&D investment → supply chain resilience”. Prior research has established that operational efficiency and R&D investment are determinants of firms’ resource allocation, risk resistance capabilities, and stable supply chain operations—thereby directly influencing supply chain resilience. The pressure for rectification and the resource crowding-out effects induced by the penalty reduce operational efficiency and constrain R&D investment space. By integrating these two mechanisms—operational efficiency and R&D investment—into a unified framework, this study systematically clarifies the transmission pathways through which environmental penalties affect corporate supply chain resilience. It therefore provides new theoretical insights and empirical evidence to help firms respond to environmental regulation shocks and strengthen their supply chain risk resistance.

Third, this paper provides trade-off insights and empirical evidence for optimizing environmental regulation policies. It shows that environmental penalties not only improve environmental governance performance but also cause negative economic effects on supply chain resilience through channels such as operational efficiency and R&D investment. These results suggest a potential trade-off between the environmental benefits of regulation and the stability of micro-level supply chains. Accordingly, the findings help policymakers to consider environmental objectives, business stability, and industrial and supply chain security when designing and implementing environmental penalties. They also offer empirical support for policy refinements such as differentiated regulation, flexible enforcement, and rectification incentives, thereby enhancing the precision and comprehensiveness of environmental governance policies.

2. Literature Review

In China’s environmental governance system, environmental penalties are an important law enforcement tool used by ecological and environmental authorities against corporate environmental violations. They are typically imposed in forms such as fines and orders for corrective action, based on laws and regulations including the Environmental Protection Law. It should be noted that existing literature, when discussing environmental penalties, often places them within the broad framework of “environmental regulation” for analysis. However, “Environmental regulation” itself is a highly composite concept, encompassing at least four levels. First, there are formal environmental policies, such as the revision of the Environmental Protection Law, carbon trading pilots, and low-carbon city pilots. These policies are macro-level and comprehensive in nature. They are typically incorporated into difference-in-differences models as quasi-natural experiments. Second, there is regulatory intensity. It is often measured by indicators such as regional-level pollutant emission standards, the frequency of environmental protection terms, and the allocation of environmental protection personnel. This reflects the overall level of enforcement willingness and resource input. Third, there is the disclosure of environmental violation information, i.e., information that a firm’s environmental violations are publicly exposed, such as environmental administrative penalty decisions and environmental protection blacklists. Its core function lies in reputation transmission and market punishment. Fourth, there is the actual penalty amount, i.e., the monetary punishment imposed on a firm for its violation, which represents the direct economic cost borne by the firm. Environmental penalties bring direct economic costs and reputational losses, exerting a sustained impact on corporate business decisions. Building on this conceptual clarification, the literature review below reorganizes existing findings around three key controversies that are directly relevant to the relationship between environmental penalties and corporate supply chain resilience.

A central debate in the literature concerns how environmental penalties affect corporate behavior, particularly environmental investment and green innovation. One strand emphasizes the deterrence effect. Proponents argue that environmental penalties raise the cost of non-compliance and send strong regulatory signals, thereby pressuring firms to strengthen environmental governance and increase green innovation. Empirical evidence from Chinese A-share listed companies shows that environmental penalties significantly increase corporate environmental investment, and the higher the penalty amount and the higher the administrative level, the more pronounced the promoting effect (He et al., 2022) [9]. Similarly, a study on German firms finds that environmental fines can improve corporate sustainability performance, with good corporate governance and high institutional ownership amplifying this effect (Ahfeeth & Çelebi, 2025) [10]. In the realm of green innovation, the deterrence effect suggests that penalties drive peer firms to improve both the quantity and quality of green technological innovation (Chen & Zhan, 2022) [11]. An opposing strand highlights the crowding-out effect. Researchers contend that environmental penalties directly consume financial resources, thereby crowding out R&D investment, and particularly inhibiting substantive green innovation (Zhang et al., 2025) [12]. From the perspective of challenge-threat theory, Peng & Ma (2025) [13] advance this debate to a nonlinear level. Using Chinese A-share listed companies from 2008 to 2020, they find that environmental penalties promote digitalization by reducing cash flow volatility on one hand, but hinder digitalization by inhibiting long-term investment on the other, resulting in an inverted U-shaped pattern. This suggests that the deterrence and crowding-out effects may coexist, and their net effect depends on penalty intensity and firm characteristics.

A second debate concerns whether environmental penalties act merely as external shocks that harm firms, or whether they can also trigger positive internal governance adjustments. The external shock perspective emphasizes that penalties, as negative events, damage firm value through reputation transmission and financing channels. Early research reveals that environmental violations have a significant negative impact on a firm’s equity market value, and the market loss is correlated with the size of the fine (Karpoff et al., 2005) [14]. In the Chinese context, even when the fine amount is limited, the disclosure of environmental violation information still triggers negative stock price reactions, and reputational loss constitutes an important component of the economic consequences of environmental penalties (Hossain et al., 2024) [15]. Furthermore, the frequency and severity of penalties are negatively correlated with a firm’s incremental bank loans, forming a “financing penalty” through reputational damage and increased risk (Chen et al., 2025) [16]. The governance optimization perspective, in contrast, argues that external pressure can drive firms to improve internal governance. When a firm is penalized for environmental violations, it drives other firms in the same industry to increase green mergers and acquisitions, a spillover effect achieved by raising environmental awareness among the public and corporate executives (Yang & Gu, 2024) [17]. Moreover, good corporate governance and high institutional ownership can amplify the positive effect of environmental fines on corporate sustainability performance (Ahfeeth & Çelebi, 2025) [10].

The third debate lies within the supply chain resilience literature itself, where scholars define and measure resilience from two complementary yet distinct perspectives: resistance capability (ex ante robustness and adaptability) and recovery capability (ex post speed and path of returning to normal operations). The resistance perspective emphasizes the ability to maintain continuity and absorb shocks. Ponomarov and Holcomb (2009) [1] systematically define supply chain resilience as a composite ability to maintain continuity, adapt to changes, and return to a desired state under disruption scenarios. Pettit et al. (2010) [2] construct an analytical framework based on vulnerability and capability. Han et al. (2020) [18] propose a three-dimensional measurement framework (preparation, response, and recovery) that also incorporates resistance elements. Empirically, Ren et al. (2025) [19], using a sample of Chinese A-share listed firms, find that environmental regulation significantly reduces supply chain deviations by improving vertical coordination, which reflects the ex ante resistance role of regulation. The recovery capability perspective focuses on how quickly a firm can bounce back after a disruption. Using a survival analysis of EPA penalty events in the United States, Foulon & Marsat (2023) [20] draw on the natural resource-based view and find that the smaller a firm’s environmental footprint, the shorter the time needed to recover from a penalty shock, and the stronger its financial resilience. Zhang et al. (2025) [21] also show that climate policy uncertainty negatively affects corporate green total factor productivity, and supply chain resilience can buffer this impact.

Based on the above review, it is evident that existing research has accumulated relatively rich findings regarding the economic consequences of environmental penalties and the antecedents of supply chain resilience. However, regarding the core concerns of this paper, there are still some gaps that have not been fully addressed. First, representative studies such as Foulon & Marsat (2023) [20] and Peng & Ma (2025) [13] focus on financial resilience and corporate digitalization, respectively. They have not yet extended the unit of analysis to supply chain resilience. Second, regarding mediating mechanisms, existing studies have identified paths such as R&D investment. However, operational efficiency, as a mediating channel through which environmental penalties affect supply chain resilience, has not yet been directly confirmed in the existing environmental penalty literature. Its mechanism of action on supply chain resilience remains unclear. This unexplored mediating channel constitutes the theoretical contribution of this study. Third, unlike existing studies that mostly use broad environmental regulation indicators, this paper treats the actual penalty amount in environmental penalties as an exclusive theoretical shock that is quantifiable and has a clear gradient of intensity. Compared with policy dummy variables or regulatory intensity indicators, the actual penalty amount can more precisely capture the severity of economic punishment faced by a firm and the changes in its marginal effects. Based on this, this study takes A-share listed firms as the research object. It focuses on analyzing the direction and intensity of the effect of environmental penalties, as a specific environmental regulation tool, on corporate supply chain resilience. Furthermore, it examines the mediating roles played by operational efficiency and R&D investment.

3. Theoretical Analysis

In a complex and volatile market environment, a firm’s supply chain system is exposed to multiple uncertainties arising from policies, markets, and operational conditions. As China’s ecological civilization construction continues to advance, environmental regulation has become a key institutional factor shaping corporate business behavior. As an important regulatory tool, environmental penalties can weaken corporate supply chain resilience through multiple channels. First, environmental penalties directly increase operating costs. Fines and subsequent requirements for environmental rectification consume substantial financial resources, which in turn reduce firms’ investments in critical areas such as inventory management and supply chain risk prevention and control. Second, environmental penalties are often accompanied by ongoing regulatory attention and requirements for production rectification. Firms may therefore need to adjust production processes, upgrade equipment, or even suspend parts of production, which can cause production disruptions and disturb the established operating rhythm of the supply chain. This weakens the supply chain system’s ability to withstand external shocks. Overall, environmental penalties reduce supply chain resilience by increasing operating costs, constraining resource investment, and inducing production disruptions. Moreover, the larger the penalty amount, the stronger the negative effect on supply chain resilience.

Accordingly, this study proposes the following research hypothesis:

H1. 

Environmental penalties have a negative effect on corporate supply chain resilience.

This study adopts an integrated analytical framework based on resource-based theory and resource dependence theory. Resource-based theory emphasizes that the operational efficiency and innovation capacity on which a firm relies to maintain supply chain resilience are endogenous to the firm’s specific resource and capability endowments; resource dependence theory points out that external shocks, by crowding out financial resources and external pressures, weaken a firm’s ability to stably obtain resources from capital markets. As an exogenous negative shock, environmental penalties both directly reduce a firm’s operational efficiency and inhibit R&D investment. These two pathways jointly weaken supply chain resilience.

Specifically, environmental penalties first reduce a firm’s operational efficiency. The rectification requirements accompanying a penalty directly disrupt a firm’s production plan. Upgrading environmental facilities and adjusting production processes consume a large amount of production resources, causing production process interruptions and efficiency losses. This leads to a continuous decline in production efficiency. A decline in operational efficiency reduces the stability of order delivery and weakens a firm’s ability to respond to market fluctuations and supply chain disruptions, thereby indirectly reducing supply chain resilience. Second, environmental penalties inhibit a firm’s R&D investment. Fines and rectification expenditures crowd out financial resources that could otherwise be used for technological upgrading and process innovation. As a result, firms find it difficult to optimize production processes through R&D, improve environmental performance, or quickly adjust supply chain operation models. Insufficient R&D investment weakens a firm’s ability to adapt its supply chain from a technical perspective. The firm cannot effectively reduce the risk of environmental violations, nor can it rely on technological innovation to enhance supply chain sustainability and impact resistance. This, in turn, indirectly and negatively affects supply chain resilience.

Accordingly, this study proposes the following research hypothesis:

H2. 

Operational efficiency and R&D investment each play a negative mediating role between environmental penalties and corporate supply chain resilience.

4. Research Design

4.1. Sample Selection and Data Sources

This study uses A-share listed companies from 2009 to 2024 as the research sample. The sample data are mainly obtained from the CSMAR, Wind, and CCER databases. The raw data were processed as follows: (1) samples from the financial and insurance industries were excluded; (2) abnormal listed companies, such as ST and *ST firms, were excluded; and (3) samples with missing relevant data were excluded. This process finally yielded 51,255 firm-year observations. In addition, to mitigate the impact of outliers, all continuous variables were winsorized at the 1st and 99th percentiles.

4.2. Definition of Key Variables

4.2.1. Dependent Variable: Supply Chain Resilience ($\mathrm{R}\mathrm{e}\mathrm{s}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}$)

The dependent variable is supply chain resilience. Its measurement primarily follows Zhang and Gu (2024) [22]. The entropy weight method is employed to construct a composite supply chain resilience index. This index integrates two dimensions: supply chain resistance and supply chain recovery.

Supply chain resistance refers to the ability of a supply chain to remain stable when facing external disruptions. It is measured using two indicators. First, the ratio of accounts receivable to operating revenue is used to capture the extent to which customers occupy the firm’s capital ($\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{i}\mathrm{s}1$). A lower ratio indicates greater stability in supply chain relationships; therefore, the negative natural logarithm of this ratio is used. Second, customer stability is measured by the proportion of top-five customers that maintain cooperative relationships with the firm for two consecutive years ($\mathrm{r}\mathrm{e}\mathrm{s}\mathrm{i}\mathrm{s}\mathrm{t}2$). A higher proportion reflects more stable supply chain relationships. The calculation formulas are as follows:

$${\mathrm{resist}1}_{\mathrm{it}} = -\ln ({\displaystyle \frac{\mathrm{Accounts} \mathrm{Receivable}}{\mathrm{O}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{g} \mathrm{R}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{n}\mathrm{u}\mathrm{e}}})$$

(1)

$${\mathrm{resist}2}_{\mathrm{it}}={\displaystyle \frac{\mathrm{N}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r} \mathrm{o}\mathrm{f} \mathrm{T}\mathrm{o}\mathrm{p} \mathrm{F}\mathrm{i}\mathrm{v}\mathrm{e} \mathrm{C}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{m}\mathrm{e}\mathrm{r}\mathrm{s} \mathrm{R}\mathrm{e}\mathrm{t}\mathrm{a}\mathrm{i}\mathrm{n}\mathrm{e}\mathrm{d} \mathrm{f}\mathrm{r}\mathrm{o}\mathrm{m} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{P}\mathrm{r}\mathrm{e}\mathrm{v}\mathrm{i}\mathrm{o}\mathrm{u}\mathrm{s} \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{N}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r} \mathrm{o}\mathrm{f} \mathrm{T}\mathrm{o}\mathrm{p} \mathrm{F}\mathrm{i}\mathrm{v}\mathrm{e} \mathrm{C}\mathrm{u}\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{m}\mathrm{e}\mathrm{r}\mathrm{s} \mathrm{i}\mathrm{n} \mathrm{t}\mathrm{h}\mathrm{e} \mathrm{C}\mathrm{u}\mathrm{r}\mathrm{r}\mathrm{e}\mathrm{n}\mathrm{t} \mathrm{Y}\mathrm{e}\mathrm{a}\mathrm{r}}}$$

(2)

Supply chain recovery refers to the ability of the supply chain to return to its normal operating state after deviations caused by external shocks. Following Lu et al. (2024) [23], this study measures supply chain recovery from two dimensions. First, supply–demand deviation ($\mathrm{r}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{v}1$) is proxied by the similarity between fluctuations in production and demand over the past several years. A higher degree of alignment indicates stronger supply chain recovery. Second, excess economic performance ($\mathrm{r}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{v}2$) is measured as the extent to which a firm’s actual profitability exceeds its estimated expected level. A larger positive deviation reflects a greater recovery capacity. The calculation formula is presented below:

$${\mathrm{recov}1}_{\mathrm{it}} = -{\displaystyle \frac{\mathrm{Var}\left(\mathrm{Production}\right)}{\mathrm{Var}\left(\mathrm{Demand}\right)}}$$

(3)

where Production = Cost of Goods Sold + Ending Net Inventory−Beginning Net Inventory, and Demand is measured by Cost of Goods Sold. A value of −recov1 greater than 1 indicates that production fluctuates more than demand, signifying a supply–demand mismatch and lower recovery capacity. Hence, recov1 is a negative indicator for supply chain resilience. Following standard practice, we multiply recov1 by −1 before the entropy weighting procedure to transform it into a positive indicator. Therefore, in our final resilience index, a larger value of (−recov1) reflects stronger supply chain recovery capacity.

$$\mathrm{Perform}=\mathsf{\alpha }+{\mathsf{\beta }}_1{\mathrm{Size}}_{\mathrm{it}}+{\mathsf{\beta }}_2{\mathrm{Lev}}_{\mathrm{it}}+{\mathsf{\beta }}_3{\mathrm{Growth}}_{\mathrm{it}}+{ \mathsf{\beta }}_4{\mathrm{Age}}_{\mathrm{it}}+{\mathsf{\beta }}_5{\mathrm{Board}}_{\mathrm{it}}+{\mathsf{\mu }}_{\mathrm{i}}+{ \mathsf{\lambda }}_{\mathrm{i}}+{ \mathsf{\epsilon }}_{\mathrm{it}}$$

(4)

In this model, Perform represents firm economic performance. It is measured as the ratio of earnings before interest and taxes ($\mathrm{E}\mathrm{B}\mathrm{I}\mathrm{T}$) to the number of employees. The control variables include firm size ($\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}$), leverage ratio ($\mathrm{L}\mathrm{e}\mathrm{v}$), operating revenue growth rate ($\mathrm{G}\mathrm{r}\mathrm{o}\mathrm{w}\mathrm{t}\mathrm{h}$), firm age ($\mathrm{A}\mathrm{g}\mathrm{e}$), and board size ($\mathrm{B}\mathrm{o}\mathrm{a}\mathrm{r}\mathrm{d}$). $\mathsf{\mu }\mathrm{i}$ and $\mathsf{\lambda }\mathrm{t}$ denote firm and year fixed effects, respectively. The residual term obtained from this regression model serves as $\mathrm{r}\mathrm{e}\mathrm{c}\mathrm{o}\mathrm{v}2$, following the residual-based approach in Ali and Zhang (2015) [24]. A larger residual value indicates that the firm’s actual performance exceeds expectations by a greater margin. This, in turn, suggests a stronger recovery capacity after external shocks.

Finally, the entropy weight method is applied to synthesize the four aforementioned indicators. This process yields the composite supply chain resilience index ($\mathrm{R}\mathrm{e}\mathrm{s}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}$).

4.2.2. Independent Variable: Environmental Penalty Amount ($\mathrm{P}\mathrm{u}\mathrm{n}\_\mathrm{A}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}$)

The core explanatory variable is environmental penalty intensity, measured by the annual environmental penalty amount imposed on each firm. The original data are obtained from the “Corporate Environmental Penalties” table in the CCER Research Database on Environmental Governance of Listed Companies. This database records environmental administrative penalty information for listed firms, including penalty dates, penalty reasons, penalty authorities, and penalty amounts.

This study aggregates all environmental fines received by each firm in each year to obtain the annual penalty amount. The original penalty amount is measured in ten thousand RMB. To facilitate coefficient interpretation, the original annual penalty amount is divided by 10 to construct $\mathrm{P}\mathrm{u}\mathrm{n}\_\mathrm{A}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}$. Therefore, a one-unit increase in $\mathrm{P}\mathrm{u}\mathrm{n}\_\mathrm{A}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}$ represents an increase of RMB 100,000 in the firm’s annual environmental penalty amount. All continuous variables, including $\mathrm{P}\mathrm{u}\mathrm{n}\_\mathrm{A}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}$, are winsorized at the 1st and 99th percentiles.

It should be noted that $\mathrm{P}\mathrm{u}\mathrm{n}\_\mathrm{A}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}$ captures penalty intensity rather than penalty frequency. To avoid ambiguity, this paper consistently uses the term “environmental penalty intensity” when interpreting the results. Future research may further distinguish penalty frequency, penalty type, and penalty severity if more detailed data are available.

4.2.3. Control Variables

Following Gao et al. (2026) [25], this study includes a set of control variables that may affect corporate supply chain resilience. Specifically, the selected control variables are firm size ($\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}$), operating cash flow ($\mathrm{C}\mathrm{a}\mathrm{s}\mathrm{h}\mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w}$), leverage ratio ($\mathrm{L}\mathrm{e}\mathrm{v}$), board size ($\mathrm{B}\mathrm{o}\mathrm{a}\mathrm{r}\mathrm{d}$), CEO duality ($\mathrm{D}\mathrm{u}\mathrm{a}\mathrm{l}$), fixed asset ratio ($\mathrm{F}\mathrm{I}\mathrm{X}\mathrm{E}\mathrm{D}$), firm age ($\mathrm{F}\mathrm{i}\mathrm{r}\mathrm{m}\mathrm{A}\mathrm{g}\mathrm{e}$), proportion of independent directors ($\mathrm{I}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{p}$), revenue growth rate ($\mathrm{G}\mathrm{r}\mathrm{o}\mathrm{w}\mathrm{t}\mathrm{h}$), and state-owned enterprise status ($\mathrm{S}\mathrm{O}\mathrm{E}$).

4.3. Model Specification

To examine the effect of penalty amounts on firm-level supply chain resilience, this study employs a two-way fixed-effects model. The model specification is as follows:

$${\mathrm{Resilience}}_{\mathrm{it}} ={ \mathsf{\beta }}_0 + {\mathsf{\beta }}_1{\mathrm{Pun}\_\mathrm{Amount}}_{\mathrm{it}} + \sum {\mathsf{\beta }}_{\mathrm{i}}{\mathrm{Control}}_{\mathrm{it}} +{ \mathsf{\lambda }}_1{\mathrm{Firm}}_{\mathrm{i}}+ {\mathsf{\lambda }}_2{\mathrm{Year}}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(5)

In this specification, the subscript i denotes the firm and t denotes the year. The dependent variable $\mathrm{R}\mathrm{e}\mathrm{s}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}\_\mathrm{i}\mathrm{t}$ represents the supply chain resilience of firm $\mathrm{i}$ in year $\mathrm{t}$. The key independent variable $\mathrm{P}\mathrm{u}\mathrm{n}\_\mathrm{A}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}\_\mathrm{i}\mathrm{t}$ denotes the environmental penalty intensity of firm $\mathrm{i}$ in year $\mathrm{t}$. The coefficient on $\mathrm{P}\mathrm{u}\mathrm{n}\_\mathrm{A}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}\_\mathrm{i}\mathrm{t}$ is the main parameter of interest. $\mathrm{C}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{s}\_\mathrm{i}\mathrm{t}$ represents a set of control variables, including firm size ($\mathrm{S}\mathrm{i}\mathrm{z}\mathrm{e}$), operating cash flow ($\mathrm{C}\mathrm{a}\mathrm{s}\mathrm{h}\mathrm{f}\mathrm{l}\mathrm{o}\mathrm{w}$), leverage ratio ($\mathrm{L}\mathrm{e}\mathrm{v}$), board size ($\mathrm{B}\mathrm{o}\mathrm{a}\mathrm{r}\mathrm{d}$), CEO duality $(\mathrm{D}\mathrm{u}\mathrm{a}\mathrm{l}$), fixed asset ratio ($\mathrm{F}\mathrm{I}\mathrm{X}\mathrm{E}\mathrm{D}$), firm age ($\mathrm{F}\mathrm{i}\mathrm{r}\mathrm{m}\mathrm{A}\mathrm{g}\mathrm{e}$), proportion of independent directors ($\mathrm{I}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{p}$), revenue growth rate ($\mathrm{G}\mathrm{r}\mathrm{o}\mathrm{w}\mathrm{t}\mathrm{h}$), and state-owned enterprise status ($\mathrm{S}\mathrm{O}\mathrm{E}$). Firm fixed effects and year fixed effects are included to control for time-invariant firm characteristics and common macroeconomic shocks. Standard errors are clustered at the firm level.

5. Empirical Study

5.1. Descriptive Statistics

Table 1. Descriptive Statistics of Main Variables.

	(1)	(2)	(3)	(4)	(5)	(6)
Variables	N	Mean	Std	Min	Median	Max
Resilience	51,255	0.3395	0.3456	0.0099	0.1490	0.9765
Pun_Amount	51,255	0.3238	1.4552	0.0000	0.0000	10.9900
Size	51,255	22.1668	1.3051	19.7961	21.9616	26.2497
Lev	51,255	0.4157	0.2090	0.0505	0.4058	0.9079
Cashflow	51,255	0.0456	0.0693	−0.1641	0.0453	0.2406
FIXED	51,255	0.2044	0.1562	0.0020	0.1711	0.6822
Growth	51,255	0.1417	0.3718	−0.5856	0.0900	2.1684
Board	51,255	2.2651	0.2535	1.6094	2.1972	2.8904
Indep	51,255	0.3835	0.0735	0.2500	0.3750	0.6000
FirmAge	51,255	2.9488	0.3406	1.9459	2.9957	3.6109
Dual	51,255	0.3029	0.4595	0.0000	0.0000	1.0000
SOE	51,255	0.3166	0.4652	0.0000	0.0000	1.0000

reports the descriptive statistics for the main variables. The mean value of corporate supply chain resilience ($\mathrm{R}\mathrm{e}\mathrm{s}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}$) is 0.3395, the median is 0.1490, and the standard deviation is 0.3456, indicating considerable variation in supply chain resilience among the sample firms and a slightly right-skewed overall distribution. The mean value of the environmental penalty amount ($\mathrm{P}\mathrm{u}\mathrm{n}\_\mathrm{A}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}$) is 0.3238, the median is 0, and the maximum value is 10.9900, suggesting that most firms received no environmental penalties or only small penalties, while a few firms faced relatively large penalties. This pattern reflects an uneven distribution of environmental regulatory shocks across firms.

The statistical characteristics of the other control variables, such as firm size, leverage ratio, cash flow, and fixed asset ratio, are generally consistent with those reported in existing studies. In addition, this paper conducts correlation and multicollinearity tests for the main variables. The correlation coefficient matrix shows that the correlation coefficients among the variables are generally low, with no evidence of strong linear correlations. Furthermore, the variance inflation factor (VIF) test shows that the VIF values of all explanatory variables are within a reasonable range. The maximum VIF is 4.474, and the mean VIF is 2.375, both well below the commonly used threshold, indicating that the model does not suffer from severe multicollinearity.

5.2. Baseline Regression

Table 2. Baseline regression results.

	(1)	(2)
Variables	Resilience	Resilience
Pun_Amount	0.0055 ***	−0.0026 **
	(3.2902)	(−2.4273)
Constant	0.3378 ***	−1.1889 ***
	(97.0747)	(−7.1026)
Controls	NO	YES
Firm fixed effects	NO	YES
Year fixed effects	NO	YES
Observations	51,255	51,255
R-squared	0.001	0.534

Robust t-statistics in parentheses, *** p < 0.01, ** p < 0.05.

reports the baseline regression results for the effect of environmental penalty amounts on corporate supply chain resilience. Column (1) presents a simple regression without control variables or fixed effects. The regression coefficient of the environmental penalty amount is 0.0055 and is significant at the 1% level. However, this specification does not fully control for firm heterogeneity or time-related factors, and thus its explanatory power is relatively limited.

Column (2) includes firm characteristics, governance variables, firm fixed effects, and year fixed effects. In this specification, the regression coefficient of the environmental penalty amount is −0.0026 and is significant at the 5% level. This result indicates that, after more rigorously controlling for potential confounding factors, an increase in the environmental penalty amount significantly weakens corporate supply chain resilience. This finding suggests that the baseline conclusion is based on a two-way fixed-effects identification strategy that accounts for firm-level heterogeneity and macroeconomic time shocks.

From an economic perspective, environmental penalties not only increase a firm’s one-time compliance costs but may also continuously affect the firm’s ability to cope with external shocks through disruptions to production and operational adjustments. Supply chain resilience essentially reflects a firm’s ability to maintain stable operations, repair cooperative relationships, and restore normal operations when facing uncertain shocks.

Environmental penalties often require firms to incur additional expenses within a relatively short period, including expenditures related to corrective actions, production suspensions, equipment upgrades, and administrative responses. Under strong resource constraints, these expenditures may crowd out resources that could otherwise be used for supplier collaboration, inventory optimization, and risk redundancy, thereby weakening the firm’s supply chain resilience.

Overall, the baseline regression results provide preliminary support for the core hypothesis of this paper. However, given that the significance of this negative effect is mainly at the 5% level, further validation is needed through the robustness tests and mechanism tests presented in the following sections.

5.3. Endogeneity Treatment

Although the baseline regressions include firm and year fixed effects, as well as a set of observable firm- and city-level characteristics, potential endogeneity issues still need to be carefully addressed. On the one hand, firms with weaker supply chain resilience may have relatively poor production process control and insufficient investment in pollution abatement. As a result, their environmental violations may be more easily detected and penalized, leading to higher penalty amounts and thus raising concerns about reverse causality. On the other hand, omitted time-varying factors, such as annual fluctuations in local environmental enforcement styles, may affect both penalty intensity and supply chain stability. If left unaddressed, these issues may cause the coefficient estimates from the baseline regressions to deviate from the true causal effects.

To mitigate these endogeneity concerns, this paper constructs an instrumental variable ($\mathrm{I}\mathrm{V}$) defined as the interaction between the firm’s one-year lagged environmental penalty amount and the number of thermal inversion days in the firm’s city in the current year. A two-stage least squares (2SLS) approach is then used for estimation. The logic behind this IV is as follows. Thermal inversion is an exogenous meteorological phenomenon determined by atmospheric circulation and topography. When ground-level air temperature is lower than that of the upper atmosphere, the vertical dispersion of pollutants is hindered, and air quality monitoring indicators are more likely to deteriorate. Even if a firm’s emissions remain unchanged, more inversion days increase the probability that its emissions will trigger environmental enforcement thresholds. At the same time, worsening environmental quality attracts greater public attention and higher supervisory pressure from higher-level governments, prompting local environmental protection departments to strengthen enforcement in years with frequent inversions. Firms with a history of penalties are more likely to be placed under intensified regulatory scrutiny. Therefore, this interaction term introduces relatively exogenous variation in enforcement intensity driven by thermal inversions, based on the regulatory attention captured by past penalty records. This satisfies the relevance condition for an IV.

Regarding the exclusion restriction, temperature inversion is a natural phenomenon unrelated to corporate management decisions. The one-year lagged penalty amount reflects a firm’s established risk profile in the regulatory system. Given firm fixed effects, this lagged variable mainly captures historical compliance records rather than contemporaneous strategic behavior. Thus, when an inversion shock occurs, the marginal increase in enforcement resources should be more concentrated on firms that were previously penalized, generating exogenous pressure on the current penalty amount. Conditional on both firm and year fixed effects, the effect of this interaction term on supply chain resilience should operate mainly through the institutional channel of current environmental penalties. In empirical studies in environmental and public economics, using temperature inversions as an IV for air pollution has become a widely adopted identification strategy, and its validity has been verified across various research contexts (Arceo et al., 2016; Sager, 2019) [11,26,27].

Nevertheless, it must be acknowledged that the exclusion restriction may not be fully satisfied in a strict sense. Beyond strengthening environmental enforcement and raising current penalty amounts, thermal inversions may also directly affect corporate supply chain resilience through non-penalty channels, such as production restriction policies, logistics disruptions, and public opinion pressure. Meanwhile, the one-year lagged penalty amount itself may independently affect a firm’s current production and inventory management strategies. These potential pathways could correlate the $\mathrm{I}\mathrm{V}$ with the error term, thereby violating the exclusion restriction. Therefore, the $\mathrm{I}\mathrm{V}$ estimation presented below should be interpreted as providing suggestive corroborative evidence under the stated assumptions, rather than as definitive proof of a causal mechanism. The significant second-stage coefficient, while consistent with the baseline finding, does not by itself establish the validity of the exclusion restriction, as an invalid instrument could also yield a significant estimate.

Table 3. Endogeneity test results.

	Stage 1	Stage 2
Variables	Pun_Amount	Resilience
Pun_Amount		−0.0172 ***
		(−3.4945)
IV	0.0012 ***
	(15.0963)
Constant	−2.2025 ***
	(−3.7846)
Kleibergen-Paap rk LM statistic	95.220 ***
Kleibergen-Paap rk Wald F statistic	227.903
Controls	YES	YES
Firm fixed effects	YES	YES
Year fixed effects	YES	YES
Observations	45,294	45,294

Robust t-statistics in parentheses, *** p < 0.01.

reports the 2SLS estimation results. In the first stage, the coefficient of the instrumental variable is positive and significant at the 1% level. The Kleibergen–Paap rk LM statistic is 95.220 and significant at the 1% level, indicating that the model is not underidentified. The Kleibergen–Paap rk Wald F statistic is 227.903, well above the Stock–Yogo critical value of 16.38 for a 10% relative bias level; therefore, concerns about weak instruments can be ruled out. In the second stage, after IV correction, the coefficient of $\mathrm{P}\mathrm{u}\mathrm{n}\_\mathrm{A}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}$ is −0.0172 and is significantly negative at the 1% level. This result is not only consistent in direction with the baseline regression but also larger in absolute magnitude. It suggests that the baseline regression may have underestimated the negative effect of environmental penalties on corporate supply chain resilience due to endogeneity. After exogenous variation is used for causal identification, the negative impact of environmental penalty amounts becomes more pronounced.

5.4. Robustness Tests

To verify the robustness of the baseline regression results, this study conducts a series of robustness checks using alternative methods. The results are reported in

Table 4. Robustness test results.

	(1)	(2)	(3)	(4)
Variables	Replaced Dependent Variable	Clustering by City	Clustering by Industry	Excluding Special Period Samples
	Resilience1	Resilience	Resilience	Resilience
Pun_Amount	−0.0070 ***	−0.0026 **	−0.0026 **	−0.0034 **
	(−3.6732)	(−2.5377)	(−2.1558)	(−2.5081)
Constant	−5.3229 ***	−1.1889 ***	−1.1889 ***	−1.0043 ***
	(−17.6361)	(−4.8675)	(−6.2381)	(−5.9266)
Controls	Yes	Yes	Yes	Yes
Firm fixed effects	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes
Observations	51,255	51,255	51,255	38,420
R-squared	0.867	0.534	0.534	0.534

Robust t-statistics in parentheses, *** p < 0.01, ** p < 0.05.

.

5.4.1. Replacing the Dependent Variable

The dependent variable in this paper is supply chain resilience. In the baseline regression, the supply chain resilience index is constructed from two dimensions—resistance and recovery—using the entropy weight method. To rule out the possibility that the conclusions are driven by a specific measurement approach, this study follows Yao et al. (2025) [28] and reconstructs an alternative measure of supply chain resilience ($\mathrm{R}\mathrm{e}\mathrm{s}\mathrm{i}\mathrm{l}\mathrm{i}\mathrm{e}\mathrm{n}\mathrm{c}\mathrm{e}1$) using the entropy-weighted TOPSIS model. This alternative measure is based on five dimensions: adaptive capacity, resistance capacity, recovery capacity, human capital, and government support. The results in Column (1) show that the coefficient of $\mathrm{P}\mathrm{u}\mathrm{n}\_\mathrm{A}\mathrm{m}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}$ is −0.0070, which is significantly negative at the 1% level. The absolute value of this coefficient is larger than that in the baseline regression, indicating that the negative effect of environmental penalties on supply chain resilience does not depend on a specific indicator construction method.

$${\mathrm{Resilience}1}_{\mathrm{it}}={ \mathsf{\beta }}_0+{ \mathsf{\beta }}_1{\mathrm{Pun}\_\mathrm{Amount}}_{\mathrm{it}}+\sum {\mathsf{\beta }}_{\mathrm{i}}{\mathrm{Control}}_{\mathrm{it} }+{\mathsf{\lambda }}_1{\mathrm{Firm}}_{\mathrm{i}}+{\mathsf{\lambda }}_2{\mathrm{Year}}_{\mathrm{t}}+{\mathsf{\epsilon }}_{\mathrm{it}}$$

(6)

5.4.2. Altering the Clustering Level of Standard Errors

To account for potential within-group correlations in the error term at the city or industry level and to improve the reliability of the estimation results, this paper conducts robustness checks by adjusting the clustering level of the standard errors. Specifically, Column (2) clusters standard errors at the city level, while Column (3) clusters standard errors at the industry level. The regression coefficients and significance levels of the environmental penalty amount do not change substantially. This indicates that the core conclusion remains robust after controlling for within-group correlations at different levels.

5.4.3. Excluding Special-Period Observations: The Pandemic Years 2020–2022

The COVID-19 pandemic during 2020–2022 constituted a rare exogenous shock to corporate production, operations, and supply chain systems, which may interfere with the accurate identification of the effect of environmental penalties. After excluding observations from the three-year pandemic period, the results in Column (4) show that the coefficient of the environmental penalty amount remains significantly negative. This indicates that the baseline conclusion is not driven by observations from the special pandemic period.

5.5. Heterogeneity Analysis

The baseline regression confirms the average negative effect of environmental penalty amounts on supply chain resilience. However, whether this effect exhibits systematic differences across firms, industries, and regions remains unclear. This section examines this issue from three perspectives: the firm level, the industry level, and the regional level.

5.5.1. Firm-Level Heterogeneity

Firms at different development stages may differ significantly in their sensitivity to environmental penalties. Following Dong and Yuan (2014) [29] and Li and Wu (2018) [30], this study divides the sample into three groups based on the tertiles of firm age: young firms, growing firms, and mature firms. The regression results are presented in Columns (1)–(3) of

Table 5. Heterogeneity Analysis Results.

	(1)	(2)	(3)	(4)	(5)	(6)	(7)
Variables	Young Firm Group	Growing Firm Group	Mature Firm Group	High-Tech Industry	Non-High-Tech Industry	High Environmental Regulation	Low Environmental Regulation
Pun_Amount	−0.0068 ***	−0.0019	−0.0015	−0.0006	−0.0036 ***	−0.0011	−0.0044 ***
	(−2.6438)	(−1.3605)	(−1.0489)	(−0.3381)	(−2.7273)	(−0.7099)	(−2.8547)
Constant	−0.3231	−1.5524	1.1509 *	−1.1188 ***	−0.9595 ***	−1.2338 ***	−0.9480 ***
	(−1.0443)	(−0.8225)	(1.6745)	(−4.2413)	(−4.2518)	(−5.8905)	(−4.2554)
Controls	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Firm fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Observations	18,344	17,949	14,962	22,310	28,945	25,901	25,346
R-squared	0.555	0.697	0.717	0.549	0.541	0.604	0.583
p-value (Fisher’s test)	0.036	0.498	0.021	0.057		0.058

Robust t-statistics in parentheses, *** p < 0.01, * p < 0.1. Note: The Fisher’s test p-values for columns (1) to (3) correspond respectively to the tests of coefficient differences between the young firm group and the growing firm group, between the growing firm group and the mature firm group, and between the young firm group and the mature firm group.

.

The results show that environmental penalties have a significant negative effect only on the supply chain resilience of young firms (coefficient = −0.0068, p < 0.01), while the effects on growing and mature firms are not significant. Fisher’s permutation tests further indicate that the coefficient for the young-firm group is significantly different from those for the growing-firm group (p = 0.036) and the mature-firm group (p = 0.021), whereas no significant difference exists between the latter two groups (p = 0.498).

This may be because young firms typically have more limited internal resources and less operational experience, making it harder for them to maintain operational efficiency and sustain R&D investment after a penalty, which in turn leads to a substantial decline in supply chain resilience. In contrast, growing and mature firms have accumulated stronger resource reserves and more stable supply chain partnerships. They can absorb the shock of penalties through measures such as diversified sourcing, inventory buffering, or green transformation. Therefore, the impact of environmental penalties on their supply chain resilience is not significant.

5.5.2. Industry-Level Heterogeneity

The impact of environmental penalties varies significantly across industries with different attributes. Following the 2012 industry classification standard of the China Securities Regulatory Commission and the National Key Supported High-Tech Fields, and drawing on the grouping method of Shi et al. (2020) [31], this study divides the sample into high-tech and non-high-tech industries. The regression results are presented in Columns (4) and (5) of Table 5.

The results show that environmental penalties have a significant negative effect on the supply chain resilience of firms in non-high-tech industries (coefficient = −0.0036, significant at the 1% level), while the effect on firms in high-tech industries is not significant (coefficient = −0.0006, t-value = −0.3381). Fisher’s permutation test indicates that the p-value for the difference between the two coefficients is 0.057, suggesting a marginally significant difference between the two groups.

This may be because firms in high-tech industries typically possess stronger green technology innovation capabilities and receive more government policy support, such as tax incentives and preferential green credit. This allows them to transform environmental compliance pressure more quickly into production process upgrades and supply chain optimization. Moreover, their products have high added value and strong customer loyalty, so the reputational damage from penalties is relatively limited. In contrast, firms in non-high-tech industries, which are mainly traditional A-share listed firms, face stricter environmental regulatory constraints and lack sufficient transformation capacity. After receiving an environmental penalty, they may struggle to adjust production processes in the short term, which undermines their operational efficiency and further weakens supply chain resilience.

5.5.3. Regional-Level Heterogeneity

Regional environmental regulation intensity may moderate the impact of environmental penalties on supply chain resilience. Regions with high regulatory intensity typically have more developed environmental governance systems, as well as supporting green finance and green technology policies, which can provide buffering mechanisms when firms face environmental penalties. By contrast, regions with low regulatory intensity have relatively weak institutional environments, making it easier for the negative shock of penalties to be transmitted to firms’ capital chains and supply chains.

Therefore, this study conducts a grouping analysis based on environmental regulation intensity. Regional environmental regulation intensity is measured as the ratio of completed investment in industrial pollution treatment to industrial value added. Due to the large amount of missing data at the prefecture-level city level, provincial-level data are used instead, and a very small number of observations from the Hong Kong Special Administrative Region with missing values are excluded. The sample is then divided into high- and low-environmental-regulation groups based on the annual median.

The regression results in Columns (6) and (7) of Table 5 show that, in the high-regulation group, the coefficient of environmental penalties on supply chain resilience is −0.0011 and is not significant. In the low-regulation group, the coefficient is −0.0044 and is significantly negative at the 1% level. Fisher’s permutation test yields a p-value of 0.058 for the difference between the two groups, which is slightly above 0.05, indicating that the difference is only marginally significant. These results suggest that the negative effect of environmental penalties on supply chain resilience is more pronounced in low-regulation regions, whereas it is buffered to some extent in high-regulation regions.

5.6. Mechanism Analysis

The baseline results show that environmental penalty intensity is negatively associated with corporate supply chain resilience. To further examine the possible transmission channels, this section focuses on two mechanisms: operational efficiency and R&D investment. These two channels correspond to short-term operational disruption and long-term capability reduction. Operational efficiency is measured by total asset turnover ($\mathrm{a}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\_\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{n}$), and R&D investment is measured by the ratio of R&D expenditure to total assets ($\mathrm{R}\mathrm{D}1$).

To avoid relying solely on the traditional Baron–Kenny stepwise procedure, this paper reports both stepwise regression results and bootstrap mediation tests. The stepwise regressions examine whether environmental penalties affect the mediator, whether the mediator is associated with supply chain resilience after controlling for environmental penalties, and whether the coefficient of Pun_Amount changes after the mediator is included. The bootstrap test further estimates the indirect effect and its confidence interval. Nevertheless, because mediation analysis relies on additional identification assumptions, the results are interpreted as evidence of possible transmission channels rather than as proof of a complete causal chain.

(1) 

Production and Operational Efficiency

Operational efficiency reflects a firm’s ability to organize production, allocate resources, and maintain turnover efficiency. Following the approach of Shen and Qiao (2024) [32], this study uses total asset turnover ($\mathrm{a}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\_\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{n}$), defined as operating revenue divided by total assets, to measure operational efficiency. Columns (1)–(3) of

Table 6. Mediation Mechanism Analysis Results.

	(1)	(2)	(3)	(4)	(5)	(6)
Variables	Resilience	asset_turn	Resilience	Resilience	RD1	Resilience
Pun_Amount	−0.0026 **	−0.0036 ***	−0.0025 **	−0.0026 **	−0.0001 ***	−0.0025 **
	(−2.4273)	(−2.8754)	(−2.3101)	(−2.4273)	(−3.8179)	(−2.2994)
asset_turn			0.0363 ***
			(3.5314)
RD1						1.1029 ***
						(5.3991)
Constant	−1.1889 ***	1.3866 ***	−1.2393 ***	−1.1889 ***	0.0476 ***	−1.2414 ***
	(−7.1026)	(7.1420)	(−7.4019)	(−7.1026)	(6.2094)	(−7.4087)
Controls	Yes	Yes	Yes	Yes	Yes	Yes
Firm fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Year fixed effects	Yes	Yes	Yes	Yes	Yes	Yes
Observations	51,255	51,255	51,255	51,255	51,255	51,255
R-squared	0.534	0.798	0.534	0.534	0.838	0.535

Robust t-statistics in parentheses, *** p < 0.01, ** p < 0.05.

report the results for this mechanism. Column (2) shows that the coefficient of Pun_Amount is −0.0036 and is significant at the 1% level, indicating that environmental penalty intensity significantly reduces operational efficiency (Greenstone et al., 2012) [33].

Column (3) further includes asset_turn in the supply chain resilience regression. The coefficient of asset_turn is 0.0363 and is significant at the 1% level, suggesting that higher operational efficiency is associated with stronger supply chain resilience (Ambulkar et al., 2015; Christopher & Peck, 2004) [34,35]. Meanwhile, the absolute value of the coefficient of Pun_Amount decreases from −0.0026 to −0.0025 after asset_turn is included. The bootstrap results in

Table 7. Mediation Analysis: Bootstrap Evidence.

Mechanism Variable	Indirect Effect	Bootstrap SE	Bias-Corrected 95% CI Lower	Bias-Corrected 95% CI Upper
Total asset turnover	−0.0001007	0.0000510	−0.0002313	−0.0000299
R&D investment	−0.0001383	0.0000466	−0.0002428	−0.0000604

Notes: This table reports the Bootstrap mediation test results. The number of Bootstrap replications is 1000. Bias-corrected 95% confidence intervals are reported. The confidence intervals of both indirect effects do not include zero, indicating that total asset turnover and R&D investment are significant mediating channels through which environmental penalties affect corporate supply chain resilience.

show that the indirect effect through asset_turn is −0.0001007, with a bias-corrected 95% confidence interval of [−0.0002313, −0.0000299], which does not include zero (Preacher & Hayes, 2008) [36]. These results suggest that operational efficiency is a partial mediating channel through which environmental penalties weaken supply chain resilience.

(2) 

R&D Investment

R&D investment reflects a firm’s long-term capability to improve technology, optimize processes, and respond to future uncertainty. This study uses the ratio of R&D expenditure to total assets ($\mathrm{R}\mathrm{D}1$) to measure R&D investment intensity. Columns (4)–(6) of Table 6 report the results for this mechanism. Column (5) shows that the coefficient of Pun_Amount is −0.0001 and is significant at the 1% level, indicating that environmental penalty intensity significantly reduces R&D investment.

Column (6) further includes RD1 in the supply chain resilience regression. The coefficient of RD1 is 1.1029 and is significant at the 1% level, suggesting that higher R&D investment is associated with stronger supply chain resilience. At the same time, the absolute value of the coefficient of Pun_Amount decreases from −0.0026 to −0.0025 after RD1 is included. The bootstrap results in Table 7 show that the indirect effect through RD1 is −0.0001383, with a bias-corrected 95% confidence interval of [−0.0002428, −0.0000604], which does not include zero. These results suggest that R&D investment is another partial mediating channel through which environmental penalties weaken supply chain resilience (Hall, 2002; Brown et al., 2009) [37,38].

In summary, the impact of environmental penalties on supply chain resilience is not merely a direct shock. It also operates through two pathways: “declining operational efficiency” and “contracting innovation investment.” The former primarily reflects damage to a firm’s current production organization and resource turnover capacity, while the latter reflects the weakening of the firm’s technological accumulation and adaptive capacity for future risks. The results in Table 6 and Table 7 show that, after the mediating variables are included, the absolute values of the coefficients of the environmental penalty amount decrease but remain significant, and the bootstrap tests confirm that both indirect effects are significant. This indicates that the above mechanisms are more consistent with a partial mediation pattern. Thus, the economic consequences of environmental penalties are not limited to the immediate expenditure on fines. They are further transmitted to the firm’s operational system and innovation decisions, ultimately weakening corporate supply chain resilience.

6. Discussion

6.1. Main Conclusions and Policy Implications

This paper finds that environmental administrative penalties significantly reduce corporate supply chain resilience. This conclusion remains robust after a series of robustness checks and instrumental variable estimation. This indicates that the impact of penalties is not limited to immediate fine expenditures but also extends to the supply chain level through corporate operations.

This conclusion does not imply that environmental enforcement should be relaxed. Environmental penalties are an important institutional arrangement for increasing the cost of non-compliance and constraining polluting behavior. Rather, this paper emphasizes that enforcement needs to balance deterrence with firms’ adjustment processes. If rules are unclear or expectations are unstable, penalties may amplify firms’ passive adjustment costs and further affect supply chain stability. Therefore, while maintaining enforcement intensity, improving the transparency of enforcement standards and enhancing policy predictability are important conditions for reducing incidental operational disruptions.

6.2. Mechanisms and Policy Levers

The mechanism tests show that environmental penalties reduce firms’ production and operational efficiency and inhibit R&D investment. This means that penalty shocks not only impose direct costs but also affect firms’ internal resource allocation, thereby weakening supply chain resilience.

The decline in production and operational efficiency suggests that, after a penalty, firms may need to devote more resources to remediation and compliance, thereby disrupting production organization and operational turnover. Accordingly, enforcement should not stop at the “penalty-payment” stage. Instead, remediation requirements, deadlines, and acceptance standards should be made clearer, helping firms incorporate environmental compliance into normal operations and reducing the squeeze on operational efficiency caused by concentrated remediation efforts.

The inhibition of R&D investment indicates that short-term compliance pressure may crowd out firms’ long-term capacity building. Because R&D investment is related to technological improvement and risk response capabilities, if firms reduce R&D activities after a penalty, their supply chain resilience will be further affected. From this perspective, environmental enforcement should do more to promote compliance through technological upgrades, process optimization, and green transformation, while avoiding excessive crowding out of long-term investment by penalty pressure.

It is worth noting, however, that the economic significance of these mediating channels differs from their statistical significance. As reported in Table 7, the indirect effect through total asset turnover is −0.0001007, and that through R&D investment is −0.0001383. Relative to the baseline coefficient of Pun_Amount (−0.0026), these two indirect effects account for only about 3.9% and 5.3% of the total effect, respectively. In other words, although both mechanisms are statistically detectable, their magnitudes are quite small. This suggests that environmental penalties likely weaken supply chain resilience through multiple other pathways not captured here—such as reputational damage, or supplier withdrawal. It also implies that, from a policy perspective, mitigating the negative impact on supply chain resilience may require more direct interventions (e.g., preserving access to trade credit, stabilizing supplier relationships, or offering compliance-related technical assistance) rather than focusing primarily on efficiency or R&D support. The modest economic magnitude of these indirect effects does not invalidate the theoretical mechanisms, but it does caution against overemphasizing efficiency and R&D as the sole levers for policy intervention.

6.3. Heterogeneous Results and Targeted Policy

The heterogeneity analysis shows that the negative impact of environmental penalties is more pronounced among young firms, non-high-tech firms, and firms in low-regulation regions. This suggests that penalty shocks do not have the same consequences for all firms. Their impact depends on firms’ resource bases, technological capabilities, and external regulatory environments.

Young firms have limited operational experience, and their management systems and supply chain relationships are not yet fully stable. When facing a penalty, they are more prone to operational disruptions. For such firms, pure ex-post punishment has limited effectiveness. Instead, proactive compliance guidance and risk warnings are more needed so that they can incorporate environmental requirements into internal management early in their growth.

Non-high-tech firms are more significantly affected, which may be related to their relatively weak technological base and process improvement capabilities. Policy should not focus solely on penalty intensity but should also guide these firms to improve production processes and environmental management capabilities, aligning environmental compliance with efficiency gains.

Firms in low-regulation regions experience greater shocks, indicating that the long-term regulatory environment affects firms’ compliance preparation. For these regions, it is more important to improve the continuity and predictability of enforcement and reduce concentrated shocks from sporadic regulatory campaigns, so that firms can form stable compliance expectations and maintain sustained investment.

6.4. Research Limitations and Future Directions

This paper has several limitations. First, it mainly measures penalty intensity by the amount of environmental penalties and does not further distinguish among different types of penalties. Future research could examine the differential effects of various penalty types, such as fines, rectification orders, and production suspensions, on supply chain resilience. Second, this paper uses instrumental variable methods to mitigate endogeneity, but IV estimation still relies on certain identification assumptions. Subsequent studies could use more refined policy shocks or quasi-natural experiments to further test the causal relationship. Third, this paper mainly examines mechanisms from the perspectives of production and operational efficiency and R&D investment. If data permit, future research could analyze more specific transmission pathways, such as changes in corporate compliance governance, production adjustments, and supply chain relationships.

In summary, this paper shows that environmental administrative penalties have economic consequences that extend to the supply chain level. The findings do not suggest that environmental enforcement should be weakened. Instead, they indicate that stricter enforcement should be accompanied by clearer rules, more stable expectations, and more targeted compliance guidance. By improving transparency, consistency, and policy support while maintaining deterrence, environmental governance can better reduce non-compliance without imposing excessive unintended disruptions on corporate supply chain resilience.