1. Introduction
In recent years, the global climate environment has exhibited a tendency to deteriorate, with various extreme climates frequently appearing. This occurrence poses severe challenges to human survival and development, making green development a focus of common concern worldwide [
1,
2]. As the largest developing country in the world, China has continuously optimized its industrial structure and promoted green and low-carbon transformations in the economic field. Enterprises, as the micro-subjects of economic and social operations and pollution emissions, are the backbone of achieving sustainable development [
3]. However, some enterprises aiming to maximize profits still maintain a development mode characterized by high energy consumption, high pollution, and high emissions. They even illegally discharge pollutants such as wastewater and waste gas, seriously impeding the green transformation of the economy. This situation arises because, under the institutional background of environmental decentralization, local governments often relax environmental supervision in pursuit of economic development. They indulge and shield environmental violations by enterprises, gradually becoming the “umbrella” for polluting enterprises [
4,
5]. To address the problem of central environmental policies being “distorted” due to insufficient motivation for environmental governance under the local management system, the Chinese government has established the vertical regulatory mechanism represented by central environmental inspection (CEPI). This mechanism directly supervises the environmental governance behavior of local governments at different levels and promotes enterprise green transformation.
Enterprise green transformation entails protecting the ecological environment while pursuing economic benefits. However, the characteristic of enterprises as “rational economic agents” dictates that enterprises are primarily driven by self-interest and are confronted with the challenge of effectively balancing the triple bottom line—economy, society, and environment. The lack of internal motivation for enterprises to fulfill environmental responsibilities leads to the prevalence of environmental default behavior [
6]. Therefore, government-imposed environmental supervision faces inherent gaps in oversight, and relying solely on institutional pressure and other government interventions in environmental governance models may not fully constrain enterprise environmental behavior [
7]. Environmental self-discipline can fill this gap. Enterprises as independent subjects, through self-regulation and self-restraint, can promote the development of environmental protection practices; this “self-adjustment” behavior is environmental self-discipline. ISO 14001 environmental management system certification currently stands as the most representative enterprise environmental self-discipline globally [
8]. It has been proven to reduce compliance costs for enterprises in dealing with mandatory environmental regulations, optimize the allocation of internal resources, and promote green development [
9,
10]. It can be seen that the environmental self-discipline mechanism has become an indispensable part in the context of the transformation from unidimensional governmental governance to multi-dimensional governance between the government and the public and enterprises. The environmental governance strategy of heavily polluting enterprises is influenced by the internal and external coupling of environmental self-discipline and vertical supervision, and exploring the green transformation of enterprises in the context of vertical supervision cannot overlook the power of environmental self-discipline.
Existing research on CEPI has predominantly focused on its impacts on enterprise economic performance [
11], environmental performance metrics [
12], and the behavioral adjustments of local governments [
13]. However, a critical gap remains in understanding whether and how the stringent pressure exerted by CEPI fundamentally drives enterprises towards a holistic green transformation—a strategic shift characterized by simultaneous reductions in pollution intensity and enhancements in efficiency. While CEPI’s “top-down” regulatory pressure is well recognized [
4,
14], the potential synergistic role of internal corporate initiatives, such as environmental self-discipline (e.g., voluntarily adopting ISO 14001), in amplifying this transformation effect has been overlooked. Previous studies have primarily examined the effects of CEPI or corporate responses in isolation, neglecting the interaction between external institutional pressures and internal environmental governance mechanisms. This study aims to fill this gap by addressing the following questions:
RQ1. Does the vertical supervision represented by CEPI affect the enterprises’ green transformation and what are the impact pathways?
RQ2. Does synergy exist between vertical supervision and environmental self-discipline and what are the influencing factors?
To investigate the policy impact of CEPI, which is a typical form of vertical supervision, and its combined effect with environmental self-discipline on enterprise green transformation, we chose A-share listed enterprises in heavily polluting industries in China from 2013 to 2020 as the research sample. The results show that CEPI significantly promotes enterprise green transformation and synergizes with environmental self-discipline. This conclusion holds even after robustness tests, such as changing the variable measurement method, the placebo tests, and propensity score matching. Research into the process by which CEPI pressure affects enterprises identifies three primary transmission mechanisms: executive green cognition, environmental protection investment, and green innovation efficiency. Further heterogeneity analysis indicates that vertical supervision and its synergistic effect with environmental self-discipline are more significant for enterprises in areas where government–enterprise relationship is less relevant, CEPI “looks back,” and local governments have high environmental enforcement power. The marginal contributions of the current study, compared to that of previous research, include the following: First, our research is based on CEPI, which enriches the existing evaluation perspective on the effect of CEPI by identifying the effect of vertical management on enterprise green transformation. Existing research has focused on the effects of CEPI on enterprise economic performance [
15,
16] and environmental performance [
17,
18]; nonetheless, few studies assess whether pressure from inspectors can promote enterprise green transformation. The current study focuses on whether CEPI stimulates internal environmental governance motivation among enterprises under the environmental decentralization system, providing more in-depth evidence for promoting the green development of polluting enterprises at the end of the environmental governance chain.
Second, the present study explores the impact of the joint environmental governance model of vertical supervision and environmental self-discipline on enterprise green transformation, offering which a fresh perspective for constructing a “multi-collaborative governance” ecological environment protection system. Existing research on CEPI has overlooked enterprise initiatives and solely focused on its direct impacts under the “top-down” model [
19,
20]. Our study clarifies that, under the impact of CEPI, enterprises actively engage in environmental self-discipline and that external regulation strengthens their recognition of environmental protection and promotes “down-top” corporate green transformation. Our study also confirms the coupled supportive relationship between external regulation and environmental self-discipline, which provides empirical evidence for the promotion of environmental “multiple governance”.
Third, the present study explores the mechanism underlying the effect of vertical supervision on enterprise green transformation and examines the boundary conditions of this transformation, as influenced by its compound effect with environmental self-discipline. Consequently, the “black box” of vertical supervision policy is elucidated. Existing studies either have concentrated on the governance effect of CEPI from the government perspective [
21,
22] or analyzed the environmental governance strategy adjustments of polluting enterprises facing CEPI [
18,
23]. Few scholars place government and enterprises within the same institutional framework and research scenario. Therefore, our study, framed by the “central government–local government–enterprise” model, examines how CEPI influences the environmental governance decisions made by polluting enterprises, with a particular focus on the proactive environmental responsibilities undertaken by these companies. Through this research, we provide empirical insights aimed at refining the CEPI system and fostering sustainable business practices.
The paper is structured as follows:
Section 2 introduces the institutional context, theoretical background, and the development of the hypotheses.
Section 3 outlines the research methodology.
Section 4 presents the core empirical findings, followed by an analysis of the influence mechanisms in
Section 5. In
Section 6, we provide additional results from further analysis.
Section 7 offers a summary of the study’s main findings, and
Section 8 extends these conclusions to the environmental governance challenges faced by other developing countries.
3. Method
3.1. Measures
3.1.1. Dependent Variable
The dependent variable is enterprise green transformation, measured using two dimensions: emission reduction and efficiency improvement. To quantify enterprise green transformation, this study uses the logarithm of industrial sulfur dioxide emissions per unit output (Emission) and total factor productivity (Tfp), respectively. Among these approaches, certain methods have been identified for measuring enterprise total factor productivity. These techniques include the following: fixed effects (FE) method, generalized method of moments (GMM), ordinary least squares (OLS) method, semi-parametric Levinsohn–Petrin (LP) method, and Olley–Pakes (OP) method. Compared to other methods, the LP and OP techniques can more effectively address the issue of selective bias in measuring enterprise total factor productivity [
45]. However, the OP method assumes a monotonic increasing relationship between investment and productivity, which may lead to the exclusion of samples with zero investment. Levinsohn and Petrin (2003) [
46] improved the OP method by using intermediate inputs instead of investment indicators, effectively addressing endogeneity issues caused by productivity shocks. This enhancement yields more reliable results. Additionally, the LP method exhibits lower data dependency and is particularly suitable for China’s manufacturing sector samples where intermittent investment is prevalent. Therefore, this study calculates total factor productivity by using the LP method and conducts robustness tests by employing the OP technique. The calculation formula for the LP method is as follows:
In Equation (1), subscripts i and t index enterprises and years, respectively. Total output (Y) derives from firm operating revenue. Labor input (L) is quantified using total employee compensation. Capital input (K) is computed from net fixed asset value. Intermediate input (M) represents the sum of operating costs, sales expenses, administrative expenditures, and financial outlays, net of current period depreciation and amortization charges.
3.1.2. Core Explanatory Variables
Central Environmental Protection Inspection: In January 2016, the pilot phase of CEPI was launched in Hebei Province, officially commencing its first round in July 2016 (see
Table 1). Each province had inspection teams stationed in four batches to conduct inspections. In the research sample, the pilot city implementing CEPI is designated as the processing group. If an inspection group is stationed in the province where the company is registered, the virtual variable Treat is set to 1; otherwise, Treat is set to 0. This study focuses on the impact of CEPI implementation on annual enterprise green transformation; thus, the pilot period and the first two batches of CEPI are classified as Phase 1 (2016), whereas the third and fourth batches are classified as Phase 2 (2017). During the years 2016 or 2017, the virtual variable Post is set to 1; otherwise, Post is set to 0. The explanatory variable CEPI is constructed as Treat × Post, indicating that when an inspection team is stationed in the province where the company is registered, CEPI = 1 for both the current and future periods; otherwise, CEPI = 0.
- 2.
Enterprise environmental self-discipline: ISO 14001 Environmental Management System Certification is recognized as the most reliable indicator of enterprise environmental self-discipline [
47,
48]. Therefore, we base our evaluation of enterprise environmental self-discipline on whether the enterprise has obtained the ISO 14001 environmental management system certification. Specifically, if the enterprise obtains the ISO 14001 certification in the current year, ISO 14001 is set to 1; otherwise, ISO 14001 is set to 0.
3.1.3. Control Variables
Referring to the research by Wang et al. (2021) [
18] and Zeng et al. (2023) [
23], we incorporate control variables at the levels of enterprise ownership structure, financial characteristics, and comprehensive governance, as well as industry and province characteristic variables. Detailed variable definitions are provided in the
Appendix A.
3.2. Sample and Data
To more accurately observe the implementation effects of the environmental pollution index, we select a research sample consisting of listed enterprises in heavily polluting industries from the A-share listed enterprises from 2013 to 2020. The starting point chosen is 2013, whereas the endpoint selected is 2020 for the following reasons: (1) The policy was implemented starting from 2016. For estimation purposes, the sample must be retained for a certain period before the policy implementation. (2) Considering the convening of the 18th National Congress, China’s focus on environmental issues markedly increased. Choosing 2013 as the starting point can reduce the estimation error caused by the abrupt shift in central environmental governance pressure to a certain extent. (3) The sampling interval cannot be excessively lengthy; otherwise, other strategies may interfere with the estimation results.
The research sample underwent the following treatments to ensure data validity: (1) Enterprises with abnormal trading statuses (ST, *ST, PT) were excluded. (2) Enterprises with abnormal or missing data were removed. (3) Continuous variables were winsorized at the 1% and 99% levels before descriptive statistics were conducted to eliminate outlier interference on the regression test. Last, we obtained 5542 sample observations. The data in this study were sourced from the CCER, RESET, and CSMAR databases.
3.3. Model
The following empirical models are constructed based on the research settings:
Equation (2) is used to evaluate the impact of vertical supervision on enterprise green transformation, and Equation (3) examines the combined effect of vertical supervision and enterprise environmental self-discipline on green transformation. The subscripts i, j, t, and p represent enterprise, industry, time, and province, respectively, with X representing other control variables. Considering the inconsistency in industry changes, this study controls for year-industry fixed effects (Yeart × Indj) to address unobservable factors at the industry level that vary over time, a more stringent approach than controlling solely for year (Yeart) and industry (Indj) fixed effects. Firmi represents individual fixed effects, and εi,t denotes the random disturbance term. Emissionit and Tfpit denote the pollution emission intensity and total factor productivity of enterprise i in year t, respectively. Emissionit and Tfpit are the dependent variables to quantitatively examine enterprise green transformation from the dimensions of emission reduction and efficiency improvement.
7. Discussion
7.1. Main Findings
A “multiple collaborative governance” mechanism involving the government as the leader, enterprises as the primary actors, and the public as participants must be established. In this context, the impact and transmission mechanism of environmental protection inspections on enterprise green transformation need to be investigated within the collaborative framework of vertical supervision and enterprise environmental self-discipline. This research considers CEPI as an exogenous policy shock and regards heavily polluting enterprises in cities under supervision as the research object to assess the impact of vertical supervision and its coordination with environmental self-discipline on enterprise green transformation and subsequently identifies its internal mechanism. This study reveals that environmental vertical supervision significantly promotes the green transformation of heavily polluting enterprises, and a synergistic effect exists between vertical supervision and environmental self-discipline. The mechanism analysis indicates that vertical supervision can drive enterprise green transformation by improving executive green cognition, increasing environmental protection investment, and improving green innovation efficiency. Heterogeneous characteristics are observed in certain dimensions, including government–enterprise relationships, CEPI “look back” initiatives, and the intensity of environmental law enforcement. These factors influence the synergistic effects of vertical supervision and environmental self-discipline.
7.2. Theoretical Contributions
This study makes three key theoretical contributions to the environmental governance literature: First, this study breaks away from prior isolated research on either government regulation or corporate self-discipline, empirically revealing for the first time a significant synergistic effect between external vertical supervision and internal environmental self-discipline [
4,
10]. Contrary to substitutionary perspectives, environmental self-discipline functions as an amplifying mechanism that enhances regulatory efficacy [
70]. It provides strong support for institutional theory propositions on how coercive pressures activate and enhance internal normative mechanisms and operationalizes the concept of “multi-stakeholder collaborative governance” in corporate green transformation.
Second, the research elucidates the mediating pathways through which regulatory pressure translates into green transformation outcomes. By validating executive green cognition, environmental protection investment, and green innovation efficiency as critical transmission mechanisms, it addresses a fundamental gap in prior work that emphasize regulation–performance correlations while obscuring intermediate processes [
17,
20]. The integration of cognitive theory, resource-based logic, and innovation dynamics provides a granular framework for understanding how institutional pressures catalyze sustainable strategic adaptation.
Third, this study significantly enriches institutional and political economy perspectives by demonstrating that the effectiveness of synergistic governance is constrained by key boundary conditions. The research finds that strong government business ties weaken the effect, while the synergistic effect is enhanced during CEPI “look back” inspections, and local environmental enforcement intensity is also crucial. This indicates that the effectiveness of synergistic governance highly depends on the firm’s institutional environment.
7.3. Policy and Managerial Implications
Based on the aforementioned conclusions, this study presents several implications. First, for governmental bodies, we recommend integrating top-down enforcement with bottom-up corporate self-discipline through incentive-aligned regulatory instruments. These should include tax relief, expedited permitting, and procurement preferences for ISO 14001-certified firms, coupled with central–local coordination mechanisms that embed environmental KPIs in officials’ performance evaluations. Concurrently, regulatory resources should prioritize corruption-vulnerable sectors where government–enterprise collusion subverts environmental goals while institutionalizing “look back” audits to combat formalism and ensure policy fidelity.
Then, corporate leaders must reconceptualize environmental management as strategic infrastructure rather than compliance overhead. The proactive adoption of certified environmental management systems builds organizational resilience and reputational capital, particularly when coupled with executive compensation structures tied to environmental KPIs and early investments in green innovation that leverage regulatory synergies for competitive advantage.
Finally, governments and firms should co-construct third-party accountability infrastructures featuring public environmental disclosure platforms that transparently report certifications, violations, and emissions data. Such systems empower stakeholders to differentially reward sustainable performers and sanction laggards while enabling firms to credibly signal environmental legitimacy.
8. Conclusions
This study, while rooted in the unique political context of China, provides valuable theoretical contributions that are highly relevant to environmental governance in other developing economies grappling with similar institutional challenges. Economic decentralization coupled with strong local protectionism represents a common challenge across emerging economies such as India, Brazil, and Mexico [
71,
72]. In these contexts, subnational governments frequently prioritize economic growth objectives over stringent environmental enforcement. Our findings demonstrate that vertical supervision mechanisms, exemplified by China’s CEPI, offer a viable strategy to counteract collusion between local regulators and polluting enterprises. Such mechanisms achieve this by imposing credible top-down accountability, facilitating bottom-up public scrutiny through transparent complaint systems, and enforcing meaningful political sanctions for non-compliance, thereby disrupting entrenched local interests.
Furthermore, achieving sustainable development in developing nations necessitates a hybrid governance model integrating mandatory oversight with voluntary corporate standards. Vertical supervision provides the essential foundation for regulatory compliance, yet its effectiveness is significantly amplified when combined with market-based incentives and robust public participation. This integrated approach fosters true co-governance. Ultimately, a resilient environmental governance system requires the collective agency of governments, enterprises, and citizens. Such a multi-dimensional framework transcends the limitations of purely top-down regulation and embeds sustainability principles within both corporate strategy and societal norms.
This study recognizes its limitations, which can be potentially addressed in future research. First, in addition to CEPI and ISO 14001 certification, other types of regulation in vertical environmental regulation and enterprise environmental self-regulation exist. These other types and their impact on enterprise green transformation may be investigated in future research to verify the universality of the current conclusion. Second, other techniques to measure enterprise green transformation enterprises have been identified. In future studies, the measurement system for enterprise green transformation could incorporate factors such as environmental training for employees and cultivating green awareness. Third, the impact mechanism examined in this study may not provide a sufficiently comprehensive view. Subsequent research could start with the dynamic game strategy among the central government, local government, and enterprises within the context of system pressure to analyze other potential mechanisms.