Environmental Governance Effects of Natural Resource Assets Departure Audit of Leading Cadres

Since implementing the departure audit in China, academic attention has focused on both regional and enterprise levels. At the regional level, research primarily explores whether the audits improve regional environmental quality [10-13]. For instance, Xiong et al. [14] found that the departure audit can significantly reduce carbon emissions, thus improving environmental conditions. Besides, Chen et al. [5] suggested that the audit system does not directly enhance air and water quality but exerts a long-term effect on environmental governance through significant increases in regional green innovation, encompassing all aspects of production processes, social structure, and supply chain management. Li and Li [15] argued that the departure audit could significantly improve environmental governance efficiency by promoting environmental institution-building [16-20], strengthening regulatory capabilities, and increasing investment in environmental governance. On the enterprise level, Sun et al. [21], using a DID model, examined the outcomes of corporate environmental responsibility and found that audits effectively promote corporate environmental responsibility, especially in non-state-owned enterprises and those not previously recognized for environmental excellence. Zhang and Wu [22] discovered that audits significantly enhance corporate environmental protection responsibility. Regarding spillover effects [23-25], existing studies have focused on the non-environmental governance impacts of the system on heavily polluting enterprises in pilot areas. Jiang and Sun [26] found that increased regulatory penalties for heavily polluting enterprises help curb tax avoidance behavior. Nie et al. [27] studied the impact on total factor productivity in enterprises, revealing that the audits boost total factor productivity in resource-intensive and heavily polluting publicly listed companies, thereby promoting revenue growth [28]. Quan et al. [29] found that implementing the departure audit leads to higher equity capital costs for resource-intensive and heavily polluting companies. However, political connections can mitigate this effect.

Research on the Impact of Environmental Regulation on Carbon Governance

In late 18^th-century England, industrial production began to sprout and develop rapidly. This period saw the early days of the Industrial Revolution. Naturally, Britain became one of the key drivers of global economic growth. With innovations in textiles, coal mining, and metallurgy, these advances have not only changed the face of local society but also had a profound impact on Europe and the world. Environmental problems have followed, especially global warming. Carbon emissions are the prime culprit in global warming. Scholars from Western developed countries were the first to conduct empirical research into the influence of environmental regulation on carbon emissions. With the increasing prominence of climate issues, scholars from other countries have also entered this research field. Existing literature mainly explored the “green paradox” and has formed three representative views. Namely, the majority of scholars believed that environmental regulation effectively curtailed corporate carbon emissions [30-32]. Shapiro and Walker [33] studied the U.S. manufacturing industry’s output and air pollutant emissions from 1990 to 2008, finding that the decline in air pollution emissions was mainly due to stricter environmental regulations. Han et al. [34] discovered that locally binding targets for emission reduction effectively pushed enterprises to cut carbon emissions, mainly because of resource reallocation among enterprises rather than market entry and exit. However, some scholars remain skeptical of the effectiveness of environmental regulation. Wang et al. [35] evaluated the impact of the “Three Rivers and Three Lakes” policy (namely, the Huai, Hai, and Liao Rivers and Tai, Chao, and Dianchi Lakes). Lakes and rivers, just like Mother Earth's heart and blood circulation, flow across the Earth, nourishing life and maintaining ecological balance. These key water bodies are critical for water pollution control efforts that impact the socio-economic development of nearly half of China’s provinces, as well as the life quality of enterprises’ chemical oxygen demand (COD) emissions. They found that while many small polluting enterprises exited the market, the policy did not significantly affect the COD emissions of operating enterprises. Using the DID method, Bao et al. [36] analyzed the local environmental legislation and found that they did not effectively curb local carbon emissions. By contrast, in areas with vigorous law enforcement, the government adopted a series of tough laws to cut carbon emissions. Not only did they have stringent emission and environmental standards, but they also imposed economic sanctions on polluters and encouraged them to invest in green technologies. The noticeable environmental improvement came true. Li and Shen [37], using cross-provincial industrial pollution data, studied the carbon reduction effects of different regulatory tools, concluding that pollution charges significantly impacted carbon reduction, whereas subsidy and green loan programs were less effective. Some scholars argued that the impact was uncertain. China’s current environmental regulatory tools have not shown significant effects in forcing carbon reduction through technological innovation [38, 39]. Under new technological constraints, an inverted U-shaped relationship exists between formal and informal environmental regulation and unit carbon emission intensity. Specifically, both formal and informal environmental policies may play a positive role in reducing carbon emissions at an early stage. Over time, however, the effects of these controls began to weaken as people adapted to and became accustomed to stringent environmental standards, which bottlenecked their efforts and led to a rebound in unit carbon intensity to higher levels. Furthermore, the synergistic effect of both types of regulation followed a V-shaped pattern. Yang et al. [40] extended this by examining the aforementioned inverted U-shaped relationship. While loose regulations can pressure enterprises into “innovation compensation,” improving carbon productivity, stringent regulations can “crowd out” technological innovation, hindering progress.

Comments from Literature

In conclusion, while research on carbon governance and the natural resource assets departure audit is abundant, most of the literature focuses solely on the environmental performance effects of these audits [41- 45], with only a few addressing their carbon reduction impact. In existing research, few scholars can integrate the two aspects of carbon emission and carbon sequestration into a unified analytical framework. This integration is critical to the conduct of the departure audit, as it requires us not only to understand the production and absorption of greenhouse gases but also to explore their complex and dynamic interactions. Such a comprehensive analysis would enable us to better understand the environmental impacts of climate change and the strategies needed to implement measures to mitigate and adapt to them. However, current research tends to focus on a single indicator, ignoring the more complex ecosystem processes behind global warming, which can lead to misjudgments about the real situation and thus affect the effectiveness of policymaking and resource allocation. This paper uses the pilot implementation of the departure audit system, a uniquely Chinese environmental regulation, as the research context to investigate whether and how this system affects regional carbon balance. This study uses the ratio of carbon sinks (CA) to energy consumption carbon emissions (CE) as the carbon balance metric and employs the departure audit system as a quasi-natural experimental condition. In this study, by using a multi-stage dual differential (DID) model, the policy effects and transmission mechanisms of these audits on carbon balance are discussed. We aim to provide more solid empirical evidence for achieving China’s “dual carbon” goal through further empirical data analysis and deepen our understanding of the effectiveness of these audits. Through careful data collection and rigorous statistical methods, this study attempts to reveal how such audits directly or indirectly affect the relationship between economic activities and the environment. In addition, this study will evaluate the adaptability and persistence of the audit in different industries, regions, and time dimensions to comprehensively evaluate its feasibility and impact scope.

Departure Audit of Leading Cadres on Natural Resource Assets and Carbon Balance

According to public choice theory, as rational economic agents, local government officials are driven to maximize their political accomplishments during their tenure to enhance their personal values and standing. Under the previous GDP-centric performance evaluation system, officials often prioritized short-term economic growth at the expense of environmental resources, which contradicted the tenets of sustainable development. The departure audit of leading cadres on natural resource assets serves as both a state audit policy tool and an indispensable environmental regulatory measure. Its primary function is to rectify officials’ performance-related behaviors and reshape local governments’ strategic decisions, ultimately contributing to carbon reduction, increasing carbon sinks, and fostering carbon balance. Consequently, this audit mechanism influences local governments’ decision-making regarding resource and environmental policies, encouraging officials to self-regulate, manage natural resources scientifically and rationally, and harmonize ecological civilization with economic development. These audits leverage independent and professional supervisory functions to ensure leaders fulfill their fiduciary responsibilities, implement environmental policies effectively, and appropriately allocate environmental governance funds [46]. This leads to an increase in urban green space and restoration of arable land, enhancing carbon sink capacity [47]. They also encourage local resource-intensive and heavily polluting enterprises to adopt a green development mindset and accelerate green technological innovations, promoting energy saving and emission reduction [5]. The following hypothesis is proposed:

H₁: The departure audit can promote regional carbon balance.

Natural Resource Assets Departure Audit of Leading Cadres, Green Innovation, and Carbon Balance

The analysis of the impact of government environmental regulations on green innovation is largely based on the Porter Hypothesis, which suggests that appropriate environmental regulations can positively stimulate corporate innovation, enhance production efficiency, and ultimately relieve the financial burden of environmental protection. As an empowered environmental regulatory tool, the departure audit of leading cadres on natural resource assets imposes significant pressure on resource-intensive and high-pollution enterprises, compelling them to expedite their green technology innovation and contribute to the attainment of carbon neutrality.

Specifically, when local enterprises’ high-carbon emissions are subject to government regulatory constraints, industrial enterprises must gradually adopt green production methods through technological innovation to avoid legal sanctions and financial penalties, effectively reducing industrial carbon emission intensity [48]. For example, in Tianjin, since the implementation of the departure audit in 2015, there has been a notable rise in the percentage of green innovation patents held by local enterprises. (as shown in Fig. 1). According to the Tianjin Municipal Bureau of Planning and Natural Resources, by 2020, the city’s total energy consumption had decreased by over 2 million tons of standard coal compared to 2015, and energy consumption per unit of GDP had declined by 19.1%. Additionally, the annual average PM_2.5 concentration had dropped from 70 µg/m³ at the end of the 12^th Five-Year Plan to 48 µg/m³. These figures indicate that the departure audit system has indeed played a significant role in promoting corporate green transformation and carbon emission control.

Implementing the departure audit system exerts dual regulatory effects, directly intensifying environmental compliance pressures on enterprises while concurrently reshaping the incentive structures of local governance actors, thereby generating secondary influences on corporate environmental strategic choices. Research shows that after the enactment of the departure audit, corporate environmental awareness and motivation significantly increased, leading to a substantial rise in environmental investments [49]. In the short term, enterprises facing environmental regulatory pressure may passively take measures such as production halts or output reductions to meet carbon reduction targets. However, in the long run, green technology innovations and introducing clean production equipment are more reasonable options to enhance energy efficiency and reduce CO₂ emissions [50, 51]. Furthermore, the departure audit system prompts local governments to prioritize clean industry development and offer policy incentives ‒ such as tax breaks and government subsidies ‒ to enterprises with strong green innovation capabilities [52, 53]. These measures drive companies toward adopting zero-carbon energy technologies and clean production processes ‒ key drivers of regional decarbonization [54].

Fig. 1.

Changes in the number of green innovations by enterprises in Tianjin.

At the same time, enterprises actively explore carbon sink solutions to add economic value to ecological policies and products. For instance, Rizhao City employs ultra-high-temperature aerobic fermentation technology to manage waste and build carbon sink demonstration parks. By adopting green planting and production methods, besides ecological value, tea plantations and fruit orchards have also brought economic benefits. In turn, the increase in economic benefits offers financial support to enhance carbon sink capacity. In particular, the departure audit system encourages governments to increase investments in carbon sequestration and sink enhancement technologies. In Dadonggang Township, Xishuangbanna, Yunnan Province, research on the impact of distinct iron oxides on soil organic carbon accumulation has facilitated the advancement of iron oxide-based soil carbon sequestration technology. This method leverages alternating wet-dry cycles in soil conditions to promote nanomagnetite precipitation, significantly boosting soil carbon sequestration potential. Meanwhile, ongoing green agriculture and forestry projects further enhance terrestrial carbon sinks, contributing to regional carbon balance. These practices demonstrate that the departure audit system not only fosters technological innovation in emission reduction but also promotes the construction of carbon sink capacity.

It is thus proposed Hypothesis 2.

H₂: The departure audit can promote carbon balance through green technological innovation.

Natural Resource Assets Departure Audit of Leading Cadres, Industrial Structure, and Carbon Balance

Institutional economics emphasizes the crucial role of institutions in shaping economic behavior. As a result, the formation of a region's industrial structure largely depends on the economic decisions made by the local government. Governments, driven by competitive pressures for economic growth and the motive to maximize tax revenue, often prioritize rapid GDP growth over low-carbon development and environmental governance. This results in the introduction of large-scale, capital-intensive industries to achieve quick economic growth, leading to an unbalanced industrial structure, severe environmental pollution, and increased carbon emissions [55]. Conversely, local governments that strive for high-quality economic growth with high-level environmental protection tend to have a more balanced industrial structure and foster greener, low-carbon economic development. The departure audit focuses on assessing party and government officials’ green performance and holds officials accountable for fulfilling their environmental fiduciary duties throughout their lifetime [56]. This can significantly influence leadership economic decision-making, shifting their focus from “prioritizing economic development and neglecting environmental governance” to promoting a green, low-carbon industrial structure. It encourages the collaboration between enterprises and governments to develop carbon market systems and carbon benefit-sharing ecosystems that complement each other, expanding ecological carbon sink channels and absorption capacity, thereby increasing ecosystem carbon sink increments.

Implementing this audit system creates a “crowding-out effect” on enterprises in heavily polluting and high-carbon industries [57, 58]. According to the “pollution haven hypothesis”, companies choose their location based on regional environmental regulation strength, which raises the “environmental barrier” of a region, making it difficult for high-pollution, high-carbon industries to enter the local market. Therefore, the implementation of this audit system can lead to the “exit” or “restricted entry” of such enterprises, which supports the growth of green, low-carbon industries and facilitates the optimization besides the upgrading of the industries [59]. In addition, these initiatives have helped to enhance corporate competitiveness, attract more investment, and create jobs, further cementing our leadership in the global economy.

H₃: The departure audit can facilitate carbon balance through the effect of industrial structure optimization. (as depicted in Fig. 2)

Dependent Variable: Measurement of Carbon Balance

(A) Measurement of Carbon Emissions. Carbon emissions data for China are calculated using the latest accounting standards for three-scope emissions, which are defined as follows: Scope 1 is in reference to all direct emissions within a city’s jurisdiction, pertaining to those from transportation, buildings, industrial manufacturing, changes in land use in agriculture and forestry, and waste treatment activities. Scope 2 means energy-related indirect emissions outside the city’s jurisdiction mainly come from heating, cooling, and purchased electricity to meet urban consumption needs. Scope 3 indicates other indirect emissions from city activities that occur outside the jurisdiction except those mentioned in Scope 2, such as greenhouse gas emissions outside the jurisdiction.

(B) Measurement of Carbon Sinks. The carbon substance approximately accounts for forty-five percent of one plant’s dry matter unit, which can absorb about 1.62 units of carbon dioxide. Specifically, each plant can absorb about 1.62 carbon dioxide molecules per unit of dry matter, a number crucial to global climate change’s impact. By photosynthesis, plants not only provide themselves with the necessary nutrients and energy but also significantly reduce carbon emissions from the environment. Therefore, the carbon sequestration of vegetation CA_i can be expressed as:

Where NPP_i represents the net primary productivity of vegetation per unit area in region i, and S_i represents the area of region i. The net primary productivity was estimated using the CASA model, integrating meteorological and remote sensing data [60].

(C) Measurement of Carbon Balance. Carbon balance serves as the dependent variable, which is specified as the ratio of vegetation carbon sinks (CA) to energy consumption carbon emissions (CE). This balance reflects the proportion of carbon dioxide (CE) produced by the absorption of carbon dioxide (CA) by vegetation through photosynthesis to the carbon dioxide (CE) produced by the release of carbon compounds from human activities. This ratio is a key barometer of ecosystem health, reflecting nature’s response to greenhouse gas emissions. In the context of global warming, accurate assessment and control of this ratio has become an important way to mitigate the impact. Therefore, an in-depth analysis of carbon balance is of irreplaceable importance for maintaining ecological balance, protecting biodiversity, and promoting green and low-carbon economic transition. This measure can better reflect the progress of regional carbon neutrality compared to individual carbon emission and carbon sequestration figures [61, 62]. Considering carbon emissions from human activities and the level of vegetation carbon sequestration, and referring to Han et al. and Ma et al. [63, 64], the carbon balance index (CBI) is defined as:

In Equation (3), CA_i,t represents the carbon sequestration of vegetation in region i during year t, and CE_i,t represents the carbon emissions attributed to energy consumption in region i during year t. Meanwhile, the carbon balance index can be categorized into three states:

When CBI_i,t = 1, the region is in a state of carbon balance, indicating equal carbon emissions and absorption levels.

When CBI_i,t<1, the region has excessive carbon emissions and insufficient carbon sequestration, indicating a carbon deficit.

When CBI_i,t>1, the region’s carbon emissions are within the ecosystem’s capacity, indicating a carbon surplus.

Independent Variable

The independent variable is the departure audit. In 2014, the National Audit Office arranged for provincial auditors to carry out pilot projects for the departure audit. From 2014 to 2017, a total of 180 cities carried out pilot programs: 14 in 2014, 21 in 2015, 82 in 2016, and 63 in 2017. In the pilot cities, the value of Audit will be set to 1 for the pilot year and the following years. Otherwise, the value of Audit will be set to 0.

(A) Control Variables. This study also draws on related research [65] to control for other variables that may affect carbon balance, considering aspects such as macroeconomic characteristics, population and education characteristics, and fiscal characteristics of the cities. The macroeconomic characteristics include the status of the economy, the degree of financial development, and the openness level. Population and education characteristics include population size, technological level, and education standard. Additionally, the characteristics of local governments include the level of urbanization and the degree of government intervention.

(B) Mediating Variables. The green patents proportion in the total annual patent applications assesses the green innovation indicator. This figure not only mirrors the extent of emphasis enterprises or research institutions place on innovation within the domains of environmental protection and sustainable development but also constitutes a significant dimension for gauging their technological innovation prowess and market competitiveness. The industrial structure optimization indicator is obtained by calculating the difference in shares between the tertiary and the secondary industries in GDP. This method reflects the soundness of the industrial structure, furnishing policymakers with vital information concerning how to adjust and stimulate economic growth. Precisely measuring this proportion makes it possible to evaluate whether a country or region has accomplished industrial upgrading or is still entrenched in traditional manufacturing and resource-intensive sectors. Additionally, it can disclose which fields demand more investment, thereby driving the entire economy toward a more balanced and sustainable path. The specific definitions and measurement methods of the variables are detailed in Table 1.

Difference-in-Differences (DID) Model

The phased and incremental pilot implementation of the Natural Resource Asset Departure Audit provides a natural "quasi-experiment" for this study. Employing the difference-in-differences (DID) method, this paper conducts a causal inference analysis of the relevant changes during the pilot phase of the departure audit. This approach facilitates an objective evaluation of the policy’s governance effects on regional carbon balance and its underlying mechanisms. The model is specified as follows in Equation (4):

Where Controls represent the group of control variables, among which k means the number of control variables, i, representing the city, signifies the local-level variable. Similarly, t denotes the time or duration of the measurement. ε, on the other hand, denotes the random error term, indicating any unmeasured or inherent error within the model that might introduce an inconsistent result from its predictions. The city fixed effect u is associated with the fixed impact factor for city characteristics. Likewise, the time fixed effect λ represents the fixed impact factor for time considerations.

Mediating Effect Model

To verify the mediating role of green innovation conjoined with industrial structure optimization, we adopt the mediating effect test method outlined by Chen et al. [5]. This leads to the following mediating effect models: M_it represents the mediating variable indicating green innovation and industrial structure optimization. The other notations have the same meanings as previously described.

Parallel Trend Test

A prerequisite for using the DID approach is that the experimental and control groups exhibit a common trend. This means that regions implementing the policy earlier and those implementing it later, as well as regions implementing the policy and those not implementing it, should not have systematic differences before the policy implementation, or if differences exist, they should be constant. The validity of the DID method relies heavily on the assumption of a parallel trend, meaning that if the pilot policy for the departure audit is lacking, the trend of carbon balance changes between pilot cities and others would be parallel. A parallel trend test was conducted using regression analysis (Fig. 3). Moreover, the regression results in Fig. 3 reveal that the regression coefficients were not significant before the pilot audit policy. However, the coefficients became significantly positive starting from the pilot year and continuing into the second year after implementation. This demonstrates that the DID test fulfills the parallel trend assumption. Positive effects of the departure audit on carbon balance appeared in the pilot year and continued to promote carbon balance into the second year of implementation, with the degree of promotion increasing over time.

Replacing the Dependent Variable

To secure reliable findings, the carbon ecological carrying capacity (ESC) [66] was used as an alternative variable to carbon balance (CBI). ESC represents the ratio of a city’s carbon sink share to the national total relative to its carbon emissions share of the national total. This indicator helps identify the capacity of carbon sinks to offset carbon emissions, reflecting the strength of carbon balance capability. Table 4 clarifies the regression results, considering control variables and fixed effects. The carbon ecological carrying capacity (ESC) shows a significantly positive relationship with the audit pilot (Audit) at the 1% statistical level. The conclusions remain consistent.

Fig. 3.

Parallel Trend Test.

Lag Period Test

To address potential endogeneity issues between the departure audit and carbon balance (CBI), this study conducted regressions with the essential core explanatory variable, Audit, lagged by one and two periods. The regression results in Table 5 illustrate that the audit coefficients for both lagged periods are notable at the 1% level. What’s more, this suggests that the audit promotes CBI, which is compatible with earlier findings, demonstrating that the conclusion holds water.

Placebo Test

When examining the effects of a policy over multiple years using a DID model, serial correlation can introduce bias into the estimates, leading to inaccurate results. To validate whether other unobserved factors influence the empirical results, we follow the approach of Li et al. [67], as do Ren et al. [68], conducting 1,000 random samplings and regression analyses to check if the estimated coefficients significantly differed from the baseline regression results. Certainly, in Fig. 4, the regression coefficients from the randomly assigned experimental groups are concentrated around zero. This implies that the positive effect of the audit on promoting green development is not influenced by omitted variables, confirming that the regression results align with expectations.

Fig. 4.

Placebo Test.

Endogeneity Test

Given the time lag between audit initiation and observable changes in carbon balance attributable to audits and the bidirectional causal relationship between carbon balance levels and audit enforcement intensity, this study employs an instrumental variable approach to address potential endogeneity. The instrumental variable is constructed by lagging the departure audit variable by one period, and regression analyses are conducted using 2SLS, LIML, and GMM (see Table 6). The lagged departure audit serves as a valid instrument because it is uncorrelated with the error term and strongly correlated with the current-period departure audit. The regression results show that the coefficient of departure audit pilot implementation (Audit) on the Carbon Balance Index (CBI) is positive and statistically significant at the 1% level, indicating that the departure audit of leading cadres on natural resource assets promotes carbon balance. This conclusion is evidently consistent with previous findings.

Additionally, to account for the potential impact of differences in characteristics between the treatment and control groups of cities on regression results, this study employs the Propensity Score Matching Difference-in-Differences (PSM-DID) method. Trimming the samples before regression improves the comparability between treated and control groups. The study uses control variables as covariates for caliper nearest-neighbor matching, setting the caliper at 0.001. In nearest-neighbor matching, only control group individuals within the predefined caliper range are considered the closest match to the treatment group individuals. After matching, the sample size is reduced to 1,900, and a multiple-period DID regression is performed on the matched sample. The results show that audits have a positive and significant effect on CBI at the 10% level, confirming the robustness of the findings (see Table 7).

Excluding the Influence of Other Policies

Since the 18^th National Congress in China, which highlighted the “vigorous promotion of ecological civilization”, the content of the national environmental policy framework has expanded. Expanding the national environmental policy framework mainly includes developments in new lawmaking, law revision, system construction, governance goals and tasks, management systems, and the public-action system. Among them, new law-making has filled legislative gaps in relevant fields, providing legal bases and guarantees for specific regions, specific ecological and environmental problems, and emerging environmental fields. Law revision has made legal norms more specific, enhancing their enforceability and operability. The construction of systems such as the red line, environmental impact assessment, pollutant discharge permits, and ecological compensation systems has become more robust. Governance goals and tasks such as the battle against pollution, carbon peaking and neutrality, and ecosystem protection and restoration have become clearer. Management systems such as ecological-environmental zoning control and the upgrading of the environmental monitoring system have become increasingly refined. The public-action system has been gradually established, including publicity, education, and public participation. Therefore, it is necessary to verify whether changes in government environmental governance during the study period could be affected by other environmental policies, ruling out noise from similar policies. This study considers that the Central Environmental Inspection, implemented starting in 2016, will most likely interfere with the estimated results due to overlapping timelines and policy targets with the natural resource asset audit pilot. To account for this, the model was adjusted to include Environmental Inspection (Inspect) as a control variable. If a province, autonomous region, or municipality was subject to environmental inspection in that year, Inspect was assigned a value of 1; otherwise, it was 0. Through Table 8, even after including the variable (Inspect), the audit pilot continued to promote regional carbon balance. This confirms that the main regression results are reliable.

Heterogeneity Analysis

Heterogeneity Based on Geographical Location

To examine the influence of geographical location, the study divided the samples into the developed eastern regions, the resource-rich central regions, and the vast western regions for grouped regression analysis. The eastern region enjoys several natural resource advantages, primarily abundant marine resources with huge potential for marine energy, excellent port resources, and favorable climatic conditions. In the western region, the main natural resource advantages are rich energy resources, abundant water-energy resources, concentrated mineral resources, and extensive land areas. The central region mainly features diverse mineral resources, sufficient water resources, and superior agricultural resources. Table 9 shows that the natural resource assets departure audit significantly affects carbon balance in all regions, indicating that it effectively promotes carbon balance among the various regions. By and large, the promotional effect is more potent in the western regions. Compared to the eastern region, the industrial structure in the other two is more homogenous and dominated by heavy industry, which poses greater environmental risks. This may result in local officials facing higher environmental pressure. Additionally, some central and western cities face economic growth challenges and increased uncertainty due to resource depletion and price fluctuations [69]. Thus, officials in these regions may place more emphasis on ecological issues, leading to a stronger effect of the audits on carbon balance.

Heterogeneity Based on Carbon Sink Resources

Heterogeneity analysis was also conducted based on regional carbon sink resources. According to the median value, cities were divided into groups with fewer and more carbon sink resources. Table 10 shows that leadership cadre audit and carbon balance regression coefficients are overwhelmingly positive in both groups, with a stronger effect in regions with richer carbon sink resources. This depicts that the richer the carbon sink resources, the more tremendous the effect of the audits on promoting carbon balance.

Natural resource endowments determine a city’s development model [70]. Universally speaking, coal resources are the "cornerstone" of industrial development. Cities with abundant carbon resources often develop resource dependence, leading to excessive carbon emissions and significant damage to carbon sink areas, which hinders carbon balance. Thus, such cities have more room for improved carbon reduction and sequestration and can leverage the departure audit to enhance carbon balance.

Heterogeneity Based on Institutional Environment

The institutional environment also affects the effectiveness of environmental policy implementation. Natural resource asset departure audit functions as a dynamic governance process secured by both internal and external regulations, and their governance effects vary under different external regulatory environments. Elaborately, governance mechanisms are more likely to be more effective within a strictly regulated market environment. This is because enterprises are obliged to adhere to more rigorous regulations and standards, thereby enhancing their compliance and stability. Conversely, governance mechanisms might be relatively feeble in a relatively lax regulatory framework, as adequate regulations are deficient to guarantee that enterprise behaviors comply with the established legal requirements.

Considering regional differences in the institutional environment, this study divided the sample into three groups ‒ weak, medium, and strong ‒ based on the data from the China Provincial Marketization Index Report by Wang Xiaolu et al. The regression results of model (5) for each group are in Table 11. The results show that the audit pilot is not associated with carbon balance (CBI) in a group with a weak institutional environment. In contrast, in the medium and strong institutional environment groups, the coefficients between the dependent variable and the audit pilot (Audit) are remarkably positive at the 1% level. In general, regions with a stronger institutional environment are more conducive to effective governance, with better resource allocation, a robust legal system, and high-quality development policies that provide a foundation for the departure audit.

Mechanism Testing

This study included mechanism variables in the baseline regression and observed their effect on the carbon balance index to further reveal the internal mechanisms through which the new urbanization policy affects carbon balance.

Table 12 reveals that the coefficient between the natural resource asset audit pilot and the green innovation mediator is 0.005. After including the environmental regulation variable, the coefficient between the carbon balance and the green innovation mediator is 19.839, and the coefficient between the audit pilot and the carbon balance is 2.443. This indicates that green innovation acts as a mediator in promoting carbon balance, supporting H₂.

Table 13 conveys that the coefficient between the audit pilot (Audit) and the industrial structure mediator is 0.162. After including the industrial structure mediator, the coefficient between the carbon balance and the industrial structure mediator is 0.073, and the coefficient between the audit pilot and the carbon balance is 2.670. This indicates that the industrial structure mediator intervenes in promoting carbon balance, confirming Hypothesis H₃.