Government Environmental Auditing and Synergistic Governance Outcomes: Evidence from Chinese Cities

Abstract

This study aims to explore the role of government environmental auditing in promoting China’s coordinated goals of “pollution reduction, carbon mitigation, ecological expansion, and growth.” By analyzing 1959 panel data from 227 prefecture-level cities in China between 2011 and 2022, a four-dimensional evaluation framework was constructed, and empirical testing was carried out using a double machine learning method. The results indicate that environmental auditing significantly enhances the synergy of environmental governance, mainly by raising public environmental awareness, promoting industrial clustering, and fostering green innovation. Additionally, green finance provides complementary support to this process. This effect is particularly pronounced in regions with higher levels of marketization, more developed financial technology, and greater environmental expenditure. Based on these findings, this study concludes that environmental auditing plays a crucial role in promoting China’s coordinated goals of “pollution reduction, carbon mitigation, ecological expansion, and growth.” In particular, environmental auditing demonstrates its institutional value in promoting sustainable governance, especially in developing economies.

1. Introduction

1.1. Research Background

In recent years, the world has been facing increasingly severe climate change and ecological degradation challenges. Global warming, biodiversity loss, and air and water pollution have caused systemic impacts on both human society and natural ecosystems [1]. How to manage the environment and achieve sustainable development is a critical issue faced by countries worldwide [2]. In response, international frameworks such as the United Nations’ 2030 Agenda for Sustainable Development, the European Union’s Green Deal, and the Paris Agreement have underscored the need for a systems-thinking approach to achieving synergistic gains in low-carbon transition and environmental governance [3,4]. However, tensions between economic growth and ecological protection remain widespread, and the imbalance between resource exploitation intensity and environmental carrying capacity continues to intensify [5], leaving substantial room for improvement in the coordination of ecological governance.

As the world’s second-largest economy and the largest emitter of carbon dioxide, China plays a pivotal strategic role in global environmental governance. In recent years, with the continued advancement of industrialization and urbanization, the country has experienced rising resource consumption intensity, large-scale pollutant emissions, and increasing pressure on ecosystem capacity. Regional environmental issues such as smog, water eutrophication, and land degradation have become widespread, posing serious challenges to ecological improvement [5]. In response, China has actively engaged in the global climate governance agenda, taking the lead in proposing its carbon peaking and carbon neutrality goals, advancing its nationally determined contributions, and continuously strengthening its domestic institutional framework for ecological civilization. The report of the 20th National Congress of the Communist Party of China explicitly called for the coordinated advancement of pollution reduction, carbon mitigation, environmental enhancement, and green growth, emphasizing the development of an ecologically prioritized, resource-efficient, and low-carbon modernization model.

In essence, environmental and natural resource products exhibit the characteristics of public goods and are marked by strong externalities. Coupled with weak regional ecological governance, insufficient market participation, limited fiscal support, and underdeveloped external oversight mechanisms [6,7,8,9,10] (pp. 21–29, pp. 402–413, pp. 2890–2921, pp. 699–754), the market mechanism has often failed to achieve effective resource allocation. Against this backdrop, the government must play a central role in policy guidance, institutional design, and public supervision to promote a transition toward efficient, coordinated, and sustainable environmental governance. Within this context, the government’s public fiduciary responsibility has increasingly extended into the ecological domain. As a result, government environmental auditing (GEA) has gradually evolved into a critical supervisory instrument for institutionalizing ecological civilization.

Objectively speaking, GEA is a key component of the national auditing system. It aims to ensure comprehensive supervision over managing natural resource assets and fulfilling ecological and environmental responsibilities through standardized auditing procedures [11]. Its core purpose goes beyond strengthening the oversight of ecological assets; it also promotes the institutionalization of green development by enhancing information disclosure, conducting performance evaluations, and enforcing accountability mechanisms. In practice, China’s National Audit Office incorporated resource and environmental auditing into its five-year development plan as early as 2018. Subsequently, the Third Plenary Session of the 18th Central Committee proposed implementing natural resource asset audits upon the departure of leading officials, and the 14th Five-Year Plan further emphasized accelerating the institutionalization of environmental auditing. These initiatives form key institutional guarantees for reducing pollution and green, low-carbon development. Existing empirical evidence has demonstrated that GEA can improve unidimensional ecological outcomes, such as mitigating regional pollution [11], enhancing urban green innovation [12], and contributing to high-quality economic development [13]. However, whether such audits produce synergistic effects in environmental governance, especially in the integrated framework of pollution reduction, carbon mitigation, ecological conservation, and economic growth, remains underexplored.

1.2. Research Questions and Purposes

Despite the various measures China has taken to promote ecological and environmental improvements, the contradiction between economic development and ecological protection remains prominent, and the synergy of environmental governance still needs enhancement. This paper aims to explore the role of GEA in promoting the coordinated governance of China’s goals of “pollution reduction, carbon reduction, ecological expansion, and growth.” The specific objectives include empirically testing the impact of GEA on the synergy of environmental governance; analyzing its mechanisms in pollution prevention, carbon reduction, ecological expansion, and green growth; and exploring the moderating role of green finance in government environmental auditing within coordinated governance.

China is selected as the study context for three main reasons: first, as the world’s second-largest economy and a major carbon emitter, its environmental challenges are highly representative [14]; second, its environmental auditing system is relatively mature, supported by robust data and institutional infrastructure; and third, there is significant regional heterogeneity in development levels and governance capacities, enabling meaningful analysis of synergy and heterogeneity effects.

1.3. Research Gaps

Although existing studies have shown that GEA has had a positive governance effect in unilateral environmental governance and has made a positive impact on pollution reduction, green innovation, and other areas, there is still a lack of systematic research on the role of GEA in the coordinated governance of “pollution reduction, carbon reduction, ecological expansion, and growth,” especially regarding how it promotes green development through institutionalization. Therefore, this paper fills this research gap by exploring the potential and mechanisms of GEA in enhancing the synergy of environmental governance.

1.4. Research Significance

The marginal contributions of this paper are mainly reflected in the following three aspects: First, it incorporates GEA into the research framework of environmental governance synergy, enriching the empirical research on ecological objectives and environmental policy tools; second, it identifies the policy transmission mechanisms and reveals the role of GEA in coordinated governance; finally, it explores the moderating role of green finance in GEA, providing policy support for promoting the precise allocation of green finance resources and strengthening the synergy between financial tools and ecological policies.

2. Literature Review and Theoretical Analysis

2.1. Literature Review

In recent years, scholarly interest in GEA has grown steadily. For example, Khan explored the origins of environmental auditing in the 1970s in the United States, emphasizing the role of civil society activities in driving environmental auditing [15] (pp. 810–826). In contrast, research rooted in the Chinese auditing system is more prevalent, focusing primarily on specific audit events or leadership accountability audits in natural resource management. These studies examine the economic and environmental consequences of such audits, particularly in areas such as air pollution, water pollution, and economic development.

In terms of air pollution control, existing research has shown that the implementation of natural resource audits for outgoing government officials significantly reduces the concentration of PM10 and PM2.5, as well as peak emissions of sensitive pollutants such as SO₂ [16,17] (pp. 17612–17628). Moreover, such audits contribute to reducing CO₂ emissions in audited cities [18], with carbon reduction effects exhibiting spatial spillover characteristics [16] (pp. 17612–17628). In water pollution control, studies based on performance audits of water pollution management programmes such as the “Three Rivers and Three Lakes” initiative have found that GEA significantly improve local water quality [17]. Regarding economic development, implementing environmental audits in key sectors has been shown to support upgrading industrial manufacturing structures, thereby facilitating high-quality economic growth [19] (pp. 3577–3597). These findings provide a solid foundation for recognizing the governance effects of GEA.

On the other hand, the question of achieving coordinated goals across pollution reduction, carbon mitigation, ecological conservation, and economic growth has become a central concern in academic and policy circles. Research has clarified that these four goals correspond, respectively, to pollutant abatement, carbon emissions reduction, green development, and high-quality growth. Achieving them requires a green transformation of the development model, deep environmental pollution control, and promoting harmony between humans and nature [20].

Regarding measurement, existing studies commonly adopt indicators such as synergistic pollution-carbon reduction efficiency, inclusive green growth, or green development efficiency as proxies for coordinated environmental progress. However, few efforts have been made to construct a comprehensive synergy evaluation index system or analyze its distribution characteristics across regions. As for the driving factors, prior research has emphasized the role of low-carbon city pilot policies, green credit availability, and environmental regulation intensity in influencing these four goals [21,22] (pp. 376–387). Nevertheless, limited attention has been paid to supervisory mechanisms—especially the governance impacts of GEA as an institutional tool.

In summary, several key gaps remain while a relatively strong theoretical and empirical foundation has been established for studying pollution reduction, carbon mitigation, ecological enhancement, and green growth. First, existing research on coordinated growth is fragmented, and current indicator systems are typically constructed around green economy, circular economy, or low-carbon economy concepts, lacking a comprehensive framework to capture the interlinked dynamics among pollution reduction, carbon mitigation, ecological improvement, and economic growth. Second, most studies focus on metropolitan regions or specific industries at the macro level, overlooking intra-provincial city-level coordination and underlying mechanisms. Finally, although environmental regulation has been confirmed as an effective driver of synergistic outcomes, research on GEA—an essential institutional guarantee for ecological civilization—remains insufficient. Few studies have systematically evaluated its effectiveness in achieving coordinated ecological governance.

In light of these gaps, this study proposes incorporating pollution reduction, carbon mitigation, ecological enhancement, and economic growth into a unified analytical framework. It constructs a systematic synergy evaluation index system and investigates the role and mechanisms of GEA in promoting integrated and effective environmental governance.

2.2. Theoretical Analysis

According to consistency theory, a lack of coordination among multiple policy objectives may lead to policy conflicts, resource misallocation, and inefficient governance. In contrast, maintaining consistency among system-level goals can enhance the systematic implementation, stability, and overall effectiveness of policy ex [23] (pp. 23–40). In the context of China’s ecological civilization construction, the four goals, carbon mitigation, pollution reduction, ecological enhancement, and economic growth, respectively, belong to the environmental, ecological, and economic systems. Among them, carbon mitigation and pollution reduction are aligned with the country’s “dual carbon” goals, focusing on controlling carbon and pollutant emissions [24], ecological enhancement emphasizes the diversity, stability, and sustainability of ecosystems; and economic growth reflects the core demands for a green, low-carbon, and high-quality development model. These four dimensions are highly interrelated and mutually reinforcing, forming an integrated structure that collectively serves the broader goal of sustainable development. As an essential component of the Party and State supervision system, audit-based oversight should fully leverage its institutional strengths to support and enhance synergistic environmental governance.

The “immune system theory” of national auditing views auditing institutions as safeguards for economic and social health through prevention, detection, and deterrence. Within this framework, GEA helps prevent environmental risks by identifying problems early, reveals misconduct by examining the authenticity and compliance of ecological actions, and deters environmentally harmful behaviour through institutional oversight [25] (pp. 605–615). These functions ensure the effective fulfilment of ecological responsibilities by local governments and enterprises, improve the self-regulation of ecosystems, and promote the integrated advancement of pollution reduction, carbon mitigation, environmental enhancement, and green growth [11].

From the perspective of public fiduciary responsibility, the government acts as the trustee of public and natural resource assets [26] (pp. 407–416). However, local officials often neglected environmental obligations under traditional GDP-oriented evaluation systems. GEA addresses this agency problem by enhancing transparency, reducing information asymmetry, and encouraging local governments to fulfil ecological responsibilities [27]. Notably, institutionalizing natural resource asset audits and introducing lifetime accountability mechanisms for ecological damage strengthen long-term ecological governance and reduce opportunistic behaviour.

Sustainable development theory further affirms the relevance of environmental auditing. As a supervisory tool, auditing helps monitor pollution, assess the implementation of environmental policies, and evaluate cost-effectiveness in ecological management [19]. By promoting the efficient allocation of environmental resources and improving regulatory enforcement, environmental auditing contributes to the broader goal of harmonizing economic growth with ecological protection. Based on these theoretical foundations, we propose the following hypothesis:

H1: 

GEA enhances the synergistic effects of environmental governance by jointly promoting pollution reduction, carbon mitigation, ecological enhancement, and green growth.

Public choice theory suggests that public concern over environmental issues significantly influences government policy decisions and the effectiveness of environmental governance [28]. However, in practice, information asymmetry limits the public’s ability to monitor environmental quality and government accountability, weakening external oversight. As an information intermediary, government environmental auditing (GEA) enhances environmental transparency through the public disclosure of audit results, raising ecological awareness, and stimulating public demand for environmental responsibility. By disclosing data related to ecological protection and pollution control, audit reports institutionalize channels for public participation and provide the informational basis for ecological engagement. Public disclosure can attract media and public attention, prompting citizens to exert pressure through complaints or suggestions [29] (pp. 45–72). This “pressure mechanism” strengthens government enforcement, deters bureaucratic inefficiency, and encourages local governments and enterprises to intensify pollution control and ecological restoration. Furthermore, institutional responsiveness theory holds that sustained public attention generates governance pressure, compelling governments to prioritize environmental objectives and reorient policy agendas [30] (pp. 210–222), thus fostering investment in green infrastructure, carbon control, and urban greening.

New economic geography theory posits that environmental constraints shape industrial spatial structures, encouraging green industries to cluster in regions with better environmental governance capacity [31] (pp. 483–499). In traditional urban economies, the concentration of high-pollution, high-emission industries exacerbates ecological stress. GEA helps quantify the environmental impact of industrial development and identify structural imbalances. The disclosure of audit findings can guide local authorities in targeted policy interventions, driving structural transformation. By systematically assessing urban industrial structures and resource allocation efficiency, environmental audits provide incentives for structural reforms, thus improving the coordination of pollution control, carbon reduction, ecological greening, and economic upgrading [32] (pp. 56961–56982). On one hand, auditing highlights mismatches between industrial expansion and environmental constraints, motivating governments to phase out high-pollution firms and foster green industrial agglomeration. On the other hand, green clusters promote clean energy, green employment, and ecological restoration, resulting in integrated environmental and economic benefits.

Schumpeterian innovation theory emphasizes that institutional pressure and policy intervention are key external drivers of technological innovation [33]. External pressures disrupt the equilibrium of existing production models, prompting firms to pursue innovation for survival and growth. GEA exerts such pressure by exposing governance gaps, regulatory loopholes, and implementation failures, strengthening accountability for firms and governments [34]. This pressure compels innovation in pollution control and carbon reduction technologies. On the enterprise side, audit disclosures institutionalize emissions thresholds and enforcement standards. Faced with rising regulatory risks, firms are incentivized to adopt green production processes, expand clean energy use, and invest in intelligent monitoring systems to reduce environmental liabilities. On the government side, audit feedback can inform more efficient design of green finance policies, R&D subsidies, and public research platforms. Targeted policy support lowers the cost and uncertainty of innovation, accelerates the diffusion of green technologies, and fosters a policy environment conducive to coordinated low-carbon transitions.

Based on these theoretical analyses, the following hypothesis is proposed:

H2: 

GEA enhances the synergistic effects of environmental governance by increasing public environmental concern, promoting green industrial agglomeration, and stimulating green technological innovation.

Green finance theory argues that without sufficient financial support, environmental governance policies often face implementation bottlenecks [35]. From an economic standpoint, environmental governance involves substantial positive externalities, long investment cycles, and high uncertainty, which makes it challenging to rely solely on fiscal funds or corporate capital. Although GEA improves regulatory oversight, pollution control, carbon mitigation, and ecological restoration require significant investment. Green finance bridges policy objectives and market mechanisms by mobilizing financial resources to support green transitions [36]. Instruments such as green credit, green bonds, and carbon trading alleviate financing constraints for governments and enterprises, channelling private capital into environmental projects. For example, green credit policies guide investment toward environmental objectives through preferential interest rates, while carbon trading creates market-based incentives for emission reductions.

Additionally, audit results can help financial institutions identify environmentally responsible entities, optimize financing structures, and reduce information asymmetry, thus lowering investment risk. By improving the efficiency of capital allocation, green finance enables the coordinated development of green investment, low-carbon production, and ecological infrastructure. Therefore, we propose the following hypothesis:

H3: 

Green finance positively moderates the relationship between GEA and the synergistic improvement of urban environmental governance.

3. Materials and Methods

This section outlines the research procedure, explaining the study’s goals and the relationships between its components. The study uses an unbalanced panel dataset covering 227 prefecture-level cities in China from 2011 to 2022 to examine the relationship between GEA and urban environmental governance. The main goal is to explore how GEA impacts the level of environmental governance through various mechanisms, including public environmental concern, industrial agglomeration, and green technological innovation. The methods include defining key variables, specifying models, and applying Double Machine Learning (DML) to analyze the causal impact of GEA on environmental outcomes. The results aim to provide insights into the effectiveness of GEA in enhancing urban environmental governance in China.

3.1. Data Sources

This study uses an unbalanced panel dataset covering 227 prefecture-level cities in China from 2011 to 2022 to ensure data continuity and availability. Data on GEA are obtained from the China Audit Yearbook. Other variables are sourced from the China City Statistical Yearbook, the China Environmental Statistical Yearbook, the EPS database, and the Wind database. Missing values, which are relatively few, are supplemented using linear interpolation. In addition, all continuous variables are Winsorized at the 1st and 99th percentiles to mitigate the influence of outliers.

3.2. Variable Definitions

3.2.1. Dependent Variable

The dependent variable is the regional environmental governance level (EGL), a comprehensive index designed to evaluate the synergy of regional environmental governance from the perspective of “pollution reduction, carbon mitigation, ecological enhancement, and green growth.” To accurately assess regional environmental protection and sustainable development performance, this study constructs an evaluation index system covering four dimensions: pollution control, carbon emissions, ecological conservation, and economic growth (see

Table 1. Coordinated Environmental Governance Evaluation Index System.

Primary Dimension	Variable Name	Unit	Attribute
Pollution Reduction	Industrial SO2 emissions	10,000 tons	–
	Industrial smoke and dust emissions	10,000 tons	–
	Harmless treatment rate of household waste	%	+
	Centralized sewage treatment rate	%	+
Carbon Mitigation	CO2 emissions	10,000 tons	–
Ecological Enhancement	Area of park green space	Hectares	+
	Green coverage rate in built-up areas	%	+
	Built-up area	Square kilometres	+
	Energy consumption intensity	—	–
	Per capita regional GDP	Yuan	+
	Education expenditure level	—	+
Green Growth	Scientific development level	Science and tech expenditure/Fiscal expenditure	+
	Enterprise scale	Log value of total output of above-scale industrial enterprises/number of such enterprises	+
	Fiscal self-sufficiency ratio	Fiscal revenue/Fiscal expenditure	+

for details). The EGL score is calculated using the entropy weighting method. To ensure consistency with other evaluation metrics and facilitate interpretation, the EGL index is scaled up by 100.

3.2.2. Independent Variable

The explanatory variable in this study is Government Environmental Audit (GEA). Existing literature lacks a unified standard for measuring GEA. Some studies rely on quasi-natural experiments, such as the Opinions of the National Audit Office on Strengthening Resource and Environmental Auditing, the auditing of natural resource assets upon the departure of government officials, and environmental audits related to the “Three Rivers and Three Lakes” programme [37,38] (pp. 1213–1241), to identify the governance effects of ecological audits. Other scholars have measured the intensity of GEA by collecting data on audit projects disclosed by the National Audit Office and local audit institutions through official webs [18].

Since this study is conducted at the prefecture-level city scale, we manually collected data from the China Audit Yearbook. Based on environmental audit projects disclosed by the National Audit Office, provincial audit departments, and local auditing agencies, we construct a proxy for GEA using each city’s total number of environmental audit projects.

3.2.3. Control Variables

Following existing literature, this study includes several control variables that may influence the level of coordinated urban environmental governance. Specifically, we control for economic development level (ECO), government intervention level (GIL), transport accessibility (TA), trade openness (TOL), and digital economy development (DIG). The squared terms of these control variables are also included in the empirical analysis to capture potential nonlinear relationships among variables and reduce the risk of model misspecification. The detailed variable definitions are presented in

Table 2. Definitions of main variables.

Variable Type	Variable Name	Symbol	Definition
Dependent Variable	Environmental Governance Level	EGL	Measured based on the framework of “pollution reduction, carbon mitigation, ecological enhancement, and green growth”
Explanatory Variable	Government Environmental Audit	GEA	Manually collected from the China Audit Yearbook
Control Variables	Economic Development Level	ECO	Logarithm of city-level per capita GDP
	Government Intervention	GIL	Ratio of general public budget revenue to GDP
	Transport Accessibility	TA	Logarithm of urban passenger traffic volume
	Trade Openness	TOL	Ratio of total import and export trade to GDP
	Digital Economy Level	DIG	Peking University Digital Inclusive Finance Index
Mechanism Variables	Public Environmental Concern	PEC	Annual city-level environmental pollution search index
	Industrial Agglomeration	IAL	Agglomeration level of manufacturing and producer services in the city
	Green Technological Innovation (Quantity)	GTI_1	Logarithm of the number of green invention patent applications
	Green Technological Innovation (Quality)	GTI_2	Logarithm of the number of green invention patents granted
	Green Finance	GF	Calculated using entropy weighting method

.

3.2.4. Mechanism Variables

Based on the preceding theoretical analysis, this study explores the synergistic effects of GEA on environmental governance by identifying the underlying mechanisms from three perspectives: public environmental concern, industrial agglomeration, and green technological innovation.

Public environmental concern is measured following Li et al. (2022) [39], using the ecological pollution search index derived from Baidu search engine data on both PC and mobile platforms. Industrial agglomeration is assessed based on the location quotient (LQ) method, following the study by Wang et al. (2023) to capture the spatial concentration of manufacturing and producer services [40] (pp. 1061–1621). Green technological innovation is represented by the natural logarithm of the number of green invention patent applications and granted patents at the city level, reflecting the quantity and quality of innovation.

In addition, this study draws on the approach of Zhou et al. (2025) [41] to construct a Green Finance Development Index based on perspectives such as green credit, green investment, green insurance, green bonds, green support, green funds, and green equity. It further explores the complementary effects between the level of green finance and government environmental auditing in environmental governance outcomes. The indicator descriptions are shown in

Table 3. Green Finance Development Index.

Variable Type	Variable Name	Variable Measurement	Specific Calculation Method
Green Finance	Green Credit	Proportion of credit for environmental protection projects	Total credit for environmental protection projects in the city/Total credit in the province
	Green Investment	Proportion of environmental pollution control investment in GDP	Environmental pollution control investment/GDP
	Green Insurance	Degree of environmental pollution liability insurance promotion	Environmental pollution liability insurance income/Total premium income
	Green Bonds	Development level of green bonds	Total amount of green bond issuance/Total amount of all bond issuance
	Green Support	Proportion of fiscal environmental protection expenditure	Fiscal environmental protection expenditure/Total general budget expenditure
	Green Funds	Proportion of green funds	Total market value of green funds/Total market value of all funds
	Green Equity	Depth of green equity development	Carbon trading, energy rights trading, emission rights trading/Total equity market trading volume

.

In addition, to examine the complementary role of green finance, this study includes a dummy variable indicating whether a city has been designated as a pilot or demonstration zone for green finance, thereby testing the interaction between financial support and GEA in shaping ecological governance outcomes.

Descriptive statistics for the variables are presented in

Table 4. Descriptive Statistics.

Variable	Observations	Mean	Std. Dev.	Min	Max
EGL	1959	13.5706	5.8645	7.6182	43.8684
GEA	1959	3.1363	1.4971	0	9
ECO	1959	16.8090	0.8827	14.7710	19.0085
GIL	1959	0.0781	0.0250	0.0343	0.1618
TA	1959	8.1913	1.1749	4.8978	11.0662
TOL	1959	0.1995	0.2803	0.0034	1.6643
DIG	1959	190.1231	78.8740	38.3700	333.1120
PEC	1959	25.1262	24.5448	1.5616	111.6160
IAL	1959	2.5119	0.4770	1.4111	3.7133
GTI_1	1959	2.9721	1.6710	0	7.0335
GTI_2	1959	4.3893	1.6346	1.0986	8.3411
GF	1959	0.3410	0.1020	0.0640	0.6575

.

3.3. Model Specification

This study focuses on examining the relationship between GEA and the level of urban environmental governance. Existing empirical research in this field has predominantly relied on traditional causal inference models. However, these models exhibit several limitations in practical applications. For example, the Difference-in-Differences (DID) approach requires the parallel trends assumption to be satisfied; the Synthetic Control Method (SCM) assumes the treated unit does not exhibit “extreme characteristics” and is typically applicable only in one-to-many treatment-control settings. Moreover, these models often face issues related to multicollinearity and nonlinearity in empirical estimation.

To address these limitations, scholars have increasingly turned to Double Machine Learning (DML) methods [42]. DML is capable of automatically selecting high-dimensional control variables and generating a precise and valid set of covariates, which helps alleviate the problem of multicollinearity. On the other hand, DML effectively mitigates model specification errors and accommodates nonlinear relationships [43], making it particularly suitable for analyzing environmental governance outcomes in this study.

Accordingly, this paper employs a partially linear DML model to identify the causal impact of GEA on environmental governance performance. The model is specified as follows:

$$Y_{it}={\theta }_0{D}_{it}+\widehat{g}\left(X_{it}\right)+U_{it}$$

(1)

$$E\left(U_{it}|D_{it},X_{it}\right)=0$$

(2)

$$D_{it}=m\left(X_{it}\right)+V_{it}$$

(3)

$$E\left(V_{it}|X_{it}\right)=0$$

(4)

In Equations (1)–(4), i denotes the province and t represents the year. $Y_{it}$ is the dependent variable, measuring the level of coordinated environmental governance. $D_{it}$ is the key explanatory variable, representing GEA, and ${\theta }_0$ is the coefficient of primary interest. $X_{it}$ denotes the set of high-dimensional control variables, whose functional form is estimated using machine learning algorithms. $U_{it}$ is the error term, with a conditional mean of zero. $m\left(X_{it}\right)$ represents the regression function of the treatment variable on the high-dimensional control variables, which is also estimated using a machine learning algorithm. $V_{it}$ is the error term, with a conditional mean of zero.

4. Empirical Results

4.1. Baseline Regression Results

This study employs a DML model to estimate the impact of GEA on environmental governance performance. The estimation results are presented in

Table 5. Baseline Regression Results: GEA and EGL.

	(1)	(2)
Variable	EGL	EGL
GEA	0.7352 ***	0.7509 ***
	(0.0932)	(0.0929)
First-order control variables	YES	YES
Second-order control variables	NO	YES
Time Fixed Effects	YES	YES
Province Fixed Effects	YES	YES
Observations	1959	1959

Note: *** indicates significance at the 1% level. Robust standard errors are in parentheses.

. Column (1) reports the results after controlling for province-fixed effects, time-fixed effects, and baseline control variables. The coefficient of the key explanatory variable is significantly positive at the 1% level. Column (2) further includes the squared terms of the control variables. The estimated coefficient remains significantly positive at the 1% level, with a value of 0.7509, indicating that a one-unit increase in the intensity of GEA is associated with an average improvement of 0.7509 units in urban environmental governance.

In practice, environmental audits constrain polluting behaviour through direct regulatory pressures such as fines and production restrictions and act as policy signals that incentivize local governments to improve the efficiency of ecological resource allocation. For example, after implementing natural resource asset audits for outgoing officials in 2016, many local governments significantly increased their environmental governance efforts by strengthening ecological fiscal expenditures to promote industrial solid waste utilization, reduce PM2.5 concentrations, and enhance overall ecological governance capacity. These findings provide strong support for Hypothesis 1.

4.2. Robustness Tests

This study conducts a series of robustness checks to verify the robustness of the baseline regression results, including alternative data splitting ratios, model substitutions, two-sided trimming, province-year interaction terms, exclusion of specific regions, and an instrumental variable (IV) approach.

First, alternative data splitting ratios: To reduce potential bias caused by the training-test sample ratio in the DML framework, we re-estimate the model using different K-fold splits (1:2 and 1:6) to test the stability of the results.

Second, model substitution: To avoid estimation bias caused by reliance on a specific prediction algorithm, we substitute the DML algorithm with random forest and gradient boosting methods to test the consistency of results across machine learning approaches.

Third, two-sided trimming: To mitigate the influence of extreme values, the sample data are Winsorized using 3% and 5% two-sided trimming.

Fourth, province-year interaction terms: While the baseline model controls for province and year-fixed effects to account for unobserved heterogeneity, it may still overlook region-specific time trends. Thus, we introduce interaction terms between province and year to further control for time-varying unobserved regional characteristics.

Fifth, exclusion of municipalities: Considering the distinct characteristics of centrally administered municipalities (e.g., Beijing, Shanghai, Tianjin, Chongqing) in terms of policy implementation, resource allocation, and industrial structure, we exclude these cities to test whether the results are driven by region-specific heterogeneity.

Sixth, instrumental variable approach: Although we have controlled for a rich set of covariates to address potential reverse causality, endogeneity may still arise due to omitted variable bias. Following the approach of Chernozhukov et al. (2018), we adopt an IV strategy [36]. Specifically, we use the lagged value of GEA as an instrument. This variable is correlated with current audit intensity and, as a historical variable, is plausibly exogenous concerning the current error term, thus satisfying the relevance and homogeneity conditions.

Seventh, replace with a difference-in-differences model: Considering the potential bidirectional causality between government environmental auditing and environmental governance levels, we use the audit of water pollution prevention projects conducted by the National Audit Office in 18 provinces and municipalities from October 2015 to January 2016 as an exogenous shock event. The difference-in-differences (DID) model is then used to estimate the impact of government environmental auditing on environmental governance levels.

The estimation results, reported in

Table 6. Robustness Checks: Alternative Splitting Ratios and Prediction Models.

Variable	(1)	(2)	(3)	(4)
	Alt. Split Ratio		Alt. Model
	1:2	1:6	Random Forest	Gradient Boosting
GEA	0.7380 ***	0.7523 ***	0.0695 ***	0.0644 **
	(0.0920)	(0.0931)	(0.0264)	(0.0284)
First-order control variables	YES	YES	YES	YES
Second-order control variables	YES	YES	YES	YES
Time Fixed Effects	YES	YES	YES	YES
Province Fixed Effects	YES	YES	YES	YES
Observations	1959	1959	1959	1959

Note: **, and *** indicate significance at the 5%, and 1% levels, respectively. Robust standard errors are in parentheses.

and

Table 7. Robustness Checks: Trimming, Province-Year Interaction, Municipality Exclusion, IV Estimation and DID.

	(1)	(2)	(3)	(4)	(5)	(6)
Variables	Trimming		Province-Year Interaction	Municipality Excluded	IV Estimation	DID
	3%	5%
GEA	0.6597 ***	0.5968 ***	0.7508 ***	0.7509 ***	2.0137 ***	0.3758 ***
	(0.0787)	(0.0702)	(0.0929)	(0.0929)	(0.3148)	(0.1287)
First-order control variables	YES	YES	YES	YES	YES	YES
Second-order control variables	YES	YES	YES	YES	YES	YES
Time Fixed Effects	YES	YES	YES	YES	YES	YES
Province Fixed Effects	YES	YES	YES	YES	YES	YES
Observations	1959	1959	1959	1911	1476	1959

Note: *** indicates significance at the 1% level. Robust standard errors are in parentheses.

, show that the positive effect of GEA on urban ecological governance remains statistically significant across all robustness checks. Although the absolute values of the coefficients vary slightly from the baseline estimates, the core conclusion remains unchanged, confirming the robustness of our findings.

5. Mechanism Tests and Heterogeneity Analysis

To systematically examine the mechanisms and heterogeneous effects through which GEA influences coordinated ecological governance, this study conducts an in-depth analysis of “pollution reduction, carbon mitigation, ecological enhancement, and green growth.” The mechanism tests focus on three key transmission pathways: public environmental concern, industrial agglomeration, green technological innovation, and the moderating role of green finance. The heterogeneity analysis explores the boundary conditions of policy effectiveness across different levels of marketization, financial technology development, and environmental expenditure intensity, aiming to uncover the synergistic logic among policy tools and the sources of regional variation.

5.1. Mechanism Tests

5.1.1. Enhancing Public Environmental Concern

To examine whether GEA enhances coordinated environmental governance by increasing public ecological concern, this study adopts the approach of Li et al. (2022) [44], using the environmental pollution search index from Baidu (PC and mobile) proxy for public concern and conducting regression analysis. As shown in Column (1) of

Table 8. Mechanism Tests: Public Environmental Concern, Industrial Agglomeration, Green Technological Innovation, and Green Finance Policy.

	(1)	(2)	(3)	(4)	(5)
Variables	Public Environmental Concern	IAL	GTI_1	GTI_2	EGL
GEA	3.6446 ***	0.0395 ***	0.1803 ***	0.1677 ***
	(0.4176)	(0.0072)	(0.0229)	(0.0238)
GEA × GFP					1.9828 ***
					(0.2571)
First-order control variables	YES	YES	YES	YES	YES
Second-order control variables	YES	YES	YES	YES	YES
Time Fixed Effects	YES	YES	YES	YES	YES
Province Fixed Effects	YES	YES	YES	YES	YES
Observations	1959	1959	1959	1959	1959

Note: *** indicates significance at the 1% level. Robust standard errors are in parentheses.

, the estimated coefficient for GEA is significantly positive at the 1% level. Specifically, a one-unit increase in audit intensity leads to a 3.6446-unit increase in public ecological concerns.

GEA reduces information asymmetry by disclosing enterprise pollution data and holding local governments accountable for environmental performance. This shifts the public role from passive recipients to active supervisors, empowering civic participation through information disclosure and promoting a co-governance structure involving government, market, and society. These findings confirm Hypothesis 2.

5.1.2. Promoting Industrial Agglomeration

To assess whether GEA exerts a structural effect by facilitating industrial agglomeration, we regress agglomeration level as the dependent variable. The results in Column (2) of Table 8 show that the coefficient on GEA is significantly positive at the 1% level. A one-unit increase in auditing intensity raises industrial agglomeration by an average of 0.0395 units.

This suggests that by strengthening resource and ecological regulation, environmental auditing raises the entry threshold and exit pressure for polluting industries, thereby facilitating the exit or transformation of high-emission enterprises. Under auditing pressure, local governments are more likely to implement targeted incentives such as tax breaks, green credits, and land-use support to attract environmentally friendly industries, restructure the industrial base, and increase the density of green industries—thus enabling synergy between pollution reduction, carbon mitigation, and green growth. These findings support Hypothesis 2.

5.1.3. Stimulating Green Technological Innovation

To test whether GEA fosters green innovation as a channel for improving ecological governance synergy, we use the logarithms of green invention patent applications (GTI_1) and granted patents (GTI_2) as proxies for innovation quantity and quality. Columns (3) and (4) of Table 8 show that the marginal effects of auditing on GTI_1 and GTI_2 are 0.1803 and 0.1677, respectively, both significant at the 1% level. These results indicate that GEA boosts the scale of green innovation and improves the quality and commercialization efficiency of green technologies.

Objectively, audit exposure increases the probability and penalty intensity for environmental violations, raising firms’ environmental risk costs. This incentivizes firms to engage in green technological upgrading to avoid compliance risks. Simultaneously, audit pressure prompts governments to introduce supportive measures such as R&D subsidies and green patent fast-tracking, thereby reducing institutional and financial barriers to innovation and strengthening the effect of environmental governance. These results validate Hypothesis 2.

5.1.4. Complementary Role of Green Finance

To assess whether green finance moderates the effect of GEAon environmental governance, we introduce the green finance variable and add an interaction term to the regression model. As shown in Column (5) of Table 8, the interaction term is significantly positive at the 1% level, with a coefficient of 1.9828, indicating that the marginal effect of government auditing on pollution reduction, carbon mitigation, ecological enhancement, and economic growth is significantly amplified in areas with green finance policies.

Mechanistically, green finance complements environmental auditing through capital provision, risk sharing, and policy signalling. Green credit and carbon-neutral bonds help expand funding access for environmentally friendly firms, reducing upfront investment pressure in pollution control, carbon reduction, and ecological restoration. Green insurance and carbon financial derivatives mitigate market risks associated with green innovation under regulatory pressure. Additionally, green finance certification mechanisms send credible policy signals to investors, strengthen market confidence, and guide resource allocation toward sustainable development. These findings confirm Hypothesis 3.

5.2. Heterogeneity Analysis

5.2.1. Marketization Level

The baseline empirical results confirm that GEA significantly enhances the synergy of pollution reduction, carbon mitigation, ecological enhancement, and green growth at the city level. However, prior studies suggest that the maturity of market mechanisms can influence the effectiveness of environmental policy implementation and the efficiency of resource allocation, which may affect the governance outcomes of audit policies [45] (pp. 259–274). Therefore, it is necessary to explore whether the policy effects of GEA vary across regions with different levels of marketization.

To this end, following the approach of Yu and Deng (2021), this study uses the marketization index scores published in the China Marketization Index Report as a measure of regional market development [46]. Cities are grouped into high- and low-marketization categories based on the median score, and an interaction term between GEA and marketization level is constructed for heterogeneity analysis. As shown in Column (1) of

Table 9. Heterogeneity Analysis: Marketization Level, Fintech Development, and Environmental Expenditure Intensity.

	(1)	(2)	(3)
Variables	Marketization Level	Fintech Development	Environmental Expenditure Intensity
GEA × Marketization Level	0.1084 ***
	(0.0114)
GEA × Fintech Development		0.5089 ***
		(0.0263)
GEA × Environmental Expenditure Intensity			0.1728 ***
			(0.0127)
First-order control variables	YES	YES	YES
Second-order control variables	YES	YES	YES
Time Fixed Effects	YES	YES	YES
Province Fixed Effects	YES	YES	YES
Observations	1959	1959	1959

Note: *** indicates significance at the 1% level. Robust standard errors are in parentheses.

, the coefficient of the interaction term is 0.1084. It is significantly positive at the 1% level, indicating that a higher degree of marketization amplifies the synergistic governance effects of auditing and strengthens the policy’s enforcement rigidity and governance efficacy.

This result suggests that well-developed market institutions provide a more favourable environment for the functioning of GEA. On the one hand, stronger property rights protection and more transparent market rules enhance the credibility of audit information and magnify its supervisory effect. On the other hand, a higher degree of marketization implies more efficient resource allocation and greater openness to green technologies and financial tools, which together help local governments and enterprises respond more effectively to audit pressure. Therefore, the interaction between GEA and marketization highlights the importance of institutional context: auditing alone is not sufficient, but its governance effect is significantly reinforced in regions with mature market mechanisms.

5.2.2. Fintech Development Level

With the rapid advancement of next-generation information technologies, digital tools have become increasingly embedded in public governance. Existing research indicates that financial technologies (fintech) can significantly enhance the institutional effectiveness of environmental governance by improving information processing and risk identification mechanisms (Sibt-e-Ali et al., 2022) [47] (pp. 1–23). Hence, this study further investigates whether the policy effect of GEA varies across cities with different levels of fintech development.

Drawing on prior methodologies, the number of fintech firms in a city is used as a proxy for fintech development. The regression model for heterogeneity testing includes an interaction term between GEA and the fintech level. As shown in Column (2) of Table 9, the interaction term yields a coefficient of 0.5089, which is statistically significant at the 1% level. This result suggests that higher fintech capacity can significantly enhance the synergistic effects of auditing policies in environmental governance.

This finding highlights the enabling role of digital financial infrastructure. On the one hand, fintech improves the timeliness and accuracy of ecological information disclosure, making audit results more transparent and actionable. On the other hand, fintech enhances environmental risk pricing and facilitates the integration of audit feedback into green credit allocation, insurance products, and investment strategies. In regions with stronger fintech development, local governments and enterprises are better able to respond to audit pressure through innovative financial channels, thereby reinforcing policy enforcement and governance efficiency. This implies that the integration of GEA with fintech is not merely complementary but creates a reinforcing feedback loop that magnifies the overall governance outcomes.

5.2.3. Local Environmental Expenditure Intensity

In addition, this study uses the ratio of local environmental protection expenditure to gross domestic product (GDP) as a proxy for ecological expenditure intensity. It includes its interaction with GEA to examine heterogeneity in governance outcomes. As reported in Column (3) of Table 9, the interaction term has a coefficient of 0.1728, significant at the 1% level. This indicates that greater environmental expenditure intensity reinforces the effectiveness of government auditing in promoting coordinated governance outcomes.

Stronger local environmental investment substantially contributes to building regional networks for pollution control, carbon mitigation, greening, and growth. It also strengthens the deterrent effect of auditing by supporting the implementation of follow-up policies and enhancing the capacity for policy enforcement, thus facilitating more effective coordinated environmental governance. This result underscores the importance of sustained financial commitment to environmental goals, suggesting that GEA’s effectiveness is not solely determined by the audit itself, but is significantly amplified in regions where governments actively allocate resources to environmental protection.

Furthermore, the interaction between GEA and environmental expenditure highlights a complementary relationship, where financial resources provide the necessary infrastructure to support audit findings and policy implementation. Local governments with higher ecological expenditure intensity are more likely to allocate the necessary funds for green technology innovation, pollution mitigation projects, and environmental restoration efforts, which ultimately strengthens the governance capacity and aligns with the broader goals of sustainable development. This finding also supports the argument that financial investments are critical to achieving the long-term success of environmental governance strategies.

6. Discussion

This study takes consistency theory as its analytical foundation, responding to the pressing challenges of goal misalignment and policy fragmentation in China’s current ecological and environmental governance. Under the overarching objectives of pollution reduction, carbon mitigation, ecological enhancement, and green growth, GEA is a crucial institutional tool linking governance frameworks to policy implementation. Whether it can effectively coordinate the environmental, ecological, and economic systems is a key indicator of the modernization of green governance capacity. Empirical results reveal that environmental auditing significantly enhances urban ecological governance and reinforces inter-system coordination through public participation, industrial agglomeration, and green technological innovation. These findings highlight how auditing contributes to improved policy coherence and system integration. This study thus expands the functional boundaries of government auditing in ecological governance, underscoring its institutional value and practical feasibility in achieving multi-objective coordination and building a modern environmental governance system.

Compared to existing research, this study highlights the unique role of GEA in promoting multi-objective coordinated governance. Many scholars have pointed out that one of the major challenges in ecological governance is how to effectively coordinate multiple goals, such as emission reduction, pollution control, and ecological restoration [48]. As an information disclosure tool, GEA can promote policy integration and enhance the overall effectiveness of environmental governance [49] (pp. 1–43). Compared to the experiences of other countries, China exhibits unique characteristics in implementing environmental audits. For example, in Europe, countries like Finland and Sweden have environmental auditing mechanisms that not only focus on pollution control but also promote the dual benefits of economic and ecological improvements. These experiences indicate that environmental auditing can effectively drive green innovation and industrial upgrading. However, China faces greater challenges in implementing environmental audits, particularly in the context of local governments being under significant pressure for economic growth, which often limits the coordination of policy execution. This study finds that in regions with higher levels of marketization and greater environmental expenditure, the effects of environmental auditing are more pronounced, further indicating the important role of regional differences in the effectiveness of environmental audits [50].

Moreover, with the rise in green finance and technology, the role of government environmental auditing has been further expanded. Similar to Germany and the United States, environmental auditing in these countries has promoted the development of green finance policies through feedback mechanisms, advancing low-carbon technologies and green industries. These countries’ experiences suggest that the synergy between environmental auditing and green policy tools can significantly improve governance efficiency.

Nevertheless, this study has certain limitations. The use of panel data at the prefecture level may not fully capture intra-regional heterogeneity, and the environmental audit data mainly comes from publicly available government information, which may overlook the informal policy execution by local governments. Future research could explore data at the enterprise level or more disaggregated data to further examine micro-level mechanisms and the specific impact of regional differences. With the continuous development of carbon markets and green financial tools, future studies should also focus on the role of environmental auditing in promoting the precise allocation of green finance resources. In addition, combining big data analysis with field research could effectively address the limitations of a single data source. For instance, in mechanism testing, supplementing interview data to explore the conversion logic between online public search behaviour and offline actions could help in deepening the analysis of the path of public environmental concern.

7. Conclusions

This study systematically examines the impact and mechanisms of GEA on urban environmental governance within the coordinated governance framework of “pollution reduction, carbon mitigation, ecological enhancement, and green growth.” Based on the findings, several key conclusions can be drawn.

First, GEA significantly improves the level of coordinated environmental governance in cities, playing a crucial role in advancing the goals of sustainable environmental management. It enhances synergy mainly by increasing public environmental concern, promoting industrial agglomeration, and stimulating green technological innovation. Additionally, green finance provides complementary support to these efforts. Furthermore, the effects of GEA are more pronounced in regions with higher levels of marketization, more developed fintech, and greater environmental expenditure intensity. This indicates that the effectiveness of GEA is contingent on regional economic and institutional factors.

These conclusions highlight the institutional value of government environmental auditing, particularly in promoting sustainable governance in developing economies.

Based on these findings, this paper offers the following policy recommendations:

First, strengthen the regional adaptability of GEA policies. Auditing strategies should be tailored according to local conditions such as economic development level, industrial structure, and ecological carrying capacity. A differentiated, region-specific auditing approach is essential to achieve “precision auditing.” Resource allocation should prioritize regions facing severe environmental problems and greater green transition pressure. Additionally, auditing outcomes should be linked to intergovernmental fiscal transfers and ecological compensation mechanisms to incentivize local governments to fulfil their ecological responsibilities. This will enhance audit policies’ adaptability and guiding power within the overarching objectives of “carbon mitigation, pollution reduction, ecological enhancement, and green growth.”

Second, deepen the synergy among fintech, environmental auditing, and green finance. Fintech can play a critical role in ecological information disclosure, ecological risk assessment, and the dynamic supervision of green credit. Developing a green finance decision-support system based on environmental audit data and powered by fintech tools is recommended. This would enable financial institutions such as banks, insurers, and securities firms to incorporate audit results into environmental risk pricing, forming a dual mechanism of “audit-driven regulation + financial incentives.” Meanwhile, the innovation of digital green financial instruments should be promoted to improve the efficiency of resource allocation for green technology innovation and the development of green industries.

Lastly, it promotes cross-regional collaboration in environmental auditing. Achieving the goals of “pollution reduction, carbon mitigation, ecological enhancement, and green growth” often requires coordination across regions due to the cross-boundary nature of environmental externalities and factor flows. Independent regional audits are usually insufficient to cover the complete ecological functional chain. Therefore, a cross-regional environmental auditing coordination mechanism should be established based on watershed areas, urban agglomerations, and ecological functional zones. This mechanism should include unified audit standards, data sharing, and joint enforcement efforts. In particular, for areas such as air pollution control, water resource management, and ecological space restoration, it is essential to establish transboundary auditing mechanisms and shared governance platforms to promote joint prevention and control efforts and integrated resource management—thus enhancing the systemic coherence and overall effectiveness of environmental governance.