The Impact of Market-Based Environmental Regulation on Carbon Emission Intensity: An Analysis Based on Policy Texts

Abstract

Market-based environmental regulations play a crucial role in promoting local carbon emissions reduction under the context of achieving “carbon peaking and carbon neutrality goals” goals. Previous studies have mainly focused on the emission reduction effects of single policy instruments, lacking systematic measurement of market-based environmental regulation. Based on panel data from Chinese prefecture-level cities during 2011–2020, this study constructs a market-based environmental regulation index using a policy text analysis method to empirically examine its impact on carbon emission intensity and the underlying mechanisms. The research findings are as follows: (1) The impact of market-based environmental regulations on carbon emission intensity shows an inverted U-shaped relationship, indicating increased carbon emissions in the short term but favorable carbon reduction in the long term. (2) Mechanism tests reveal that market-based environmental regulations primarily influence carbon emission intensity through enterprise green innovation, showing an inverted U-shaped relationship, suggesting that such regulations may suppress enterprise innovation vitality in the short term but effectively promote green innovation in the long term. (3) The carbon reduction effect of market-based environmental regulations varies significantly across regions with different official characteristics. This study provides an important theoretical basis and policy implications for improving market-based environmental regulation policy design and enhancing carbon reduction effectiveness.

1. Introduction

Currently, global climate change has become a major challenge facing humanity, and reducing carbon emissions and achieving low-carbon development have become common aspirations of the international community [1,2]. As a key participant in global carbon reduction, China is actively assuming its responsibility as a major power, contributing Chinese wisdom and solutions to global climate governance. To this end, China has made a solemn commitment to the international community to strive to achieve carbon peak by 2030 and carbon neutrality by 2060. The realization of this ambitious goal requires not only government policy guidance but also the effective functioning of market mechanisms [3].

Market-based environmental regulation serves as a crucial policy instrument through which governments guide corporate environmental behavior via market mechanisms. Unlike command-and-control environmental regulations that directly restrict corporate emissions, market-based approaches provide firms with greater autonomy in emission reduction through price signals and economic incentives [4], primarily including pollution charges, emissions trading, and environmental subsidies. As environmental governance policies continue to evolve, market-based regulatory instruments have developed into a diverse ecosystem of tools. The carbon emissions trading system, through quota allocation and trading mechanisms, has established market-driven constraints. Research indicates that this system has significantly reduced carbon emission intensity in pilot regions while incentivizing corporate technological innovation through price signals [5,6]. Similarly, environmental taxes and fees, as pioneering market-based instruments, have demonstrated notable advantages in improving carbon production efficiency by encouraging voluntary emission reductions through the internalization of environmental costs [7,8,9]. Furthermore, innovative financial instruments such as green credit and green bonds not only provide financial support for companies’ low-carbon transition but also optimize the allocation of environmental credit resources [10,11,12]. Concurrently, government environmental subsidies have incentivized green innovation among enterprises, contributing to the continued decline in industrial carbon intensity [13].

However, existing research exhibits three significant limitations. First, studies have largely focused on evaluating individual policy instruments in isolation. In reality, China’s market-based environmental regulations have evolved into a diverse, systematic framework. Policy instruments demonstrate both cumulative and synergistic effects, making it insufficient to examine single tools independently. More critically, most existing studies assume a linear relationship between market-based environmental regulations and carbon emissions, overlooking how initial adaptation costs and institutional imperfections during the exploratory phase may actually lead to increased carbon intensity. This oversimplification fails to accurately reflect policy effectiveness and provides inadequate guidance for policy optimization. Second, while the Porter hypothesis emphasizes that environmental regulations can achieve emission reduction targets by stimulating corporate innovation, current research demonstrates a limited understanding of the theoretical mechanisms through which market-based environmental regulations influence green innovation. The green paradox theory suggests that due to the high investment requirements, long development cycles, and significant risks associated with green innovation, combined with immature market mechanisms, enterprises may exhibit a “wait-and-see effect”, postponing innovation decisions. Conversely, the Porter hypothesis argues that appropriate environmental regulations can stimulate corporate innovation drive. However, existing research has disproportionately focused on the positive effects of market-based environmental regulations in promoting green innovation, neglecting potential negative corporate responses. This theoretical limitation impedes accurate understanding of policy transmission mechanisms. Third, under China’s distinctive cadre management system, local officials are key actors in policy implementation. Their personal characteristics—including age, educational background, tenure, and origin—may significantly influence policy outcomes through their varying interpretations and applications of market-based instruments. Nevertheless, current research has paid insufficient attention to this policy implementation heterogeneity within China’s institutional context.

To address these limitations, this study develops a comprehensive analytical framework for measuring market-based environmental regulations, beginning with environmental policy texts. The methodology encompasses several key components: First, we establish a three-tier coding system—“document number-chapter-specific provision”—to systematically identify and extract market-based policy provisions related to carbon emissions trading, environmental taxes and fees, and green finance. Second, we quantify policy intensity through a five-level scoring system based on the legal authority and administrative hierarchy of policy texts. Third, we account for the dynamic adjustments in policy stock, thereby more accurately reflecting the cumulative effects of market-based environmental regulations. Building on this framework, this study thoroughly examines the mediating role of corporate green innovation and the heterogeneous effects of officials’ characteristics. This approach provides novel theoretical perspectives and policy insights for understanding how market mechanisms can achieve China’s dual carbon goals within its unique institutional context.

2. Theoretical Foundation and Hypothesis Development

2.1. Market-Based Environmental Regulation and Carbon Emission Intensity

Market-based environmental regulations refer to governmental approaches that achieve environmental policy objectives through market mechanisms rather than direct corporate intervention [4]. These regulations primarily encompass systems such as pollution charges, emissions trading, and environmental subsidies. This regulatory approach grants enterprises greater autonomy in decision-making, effectively stimulating corporate initiative [14]. While market-based environmental regulations complement the singularity and compulsory nature of command-and-control regulations, they remain in developmental stages, and imperfect market mechanisms may even produce negative effects [15]. Consequently, market-based environmental regulations exhibit two distinct effects on carbon emission intensity: the “inertia effect” and the “incentive effect”. The “inertia effect” establishes a positive correlation between environmental regulations and corporate carbon emission intensity—as regulatory intensity increases, corporate carbon emissions paradoxically rise. Conversely, the “incentive effect” creates a negative correlation—stronger regulations lead to lower corporate carbon emission intensity.

Research based on the green paradox theory provides a cogent explanation for the “inertia effect” observed in market-based environmental regulations [16]. This effect stems from several interrelated factors: First, green innovation exhibits significant dual externality characteristics [17,18]. Environmental externality prevents firms from fully capturing the benefits of emission reductions, while knowledge externality allows innovation outcomes to be easily imitated and utilized by others, reducing proprietary returns. This dual externality creates an innovation incentive deficit. Additionally, while technological innovation requires extended research and development cycles, market-based tools (such as quota purchases and environmental tax payments) offer immediate compliance solutions. Under mild carbon price constraints, the direct costs of market transactions are substantially lower than green technology R&D investments, further diminishing corporate innovation motivation. Second, systemic deficiencies exist in market mechanisms during the initial regulatory phase. The carbon emissions trading system has limited industry coverage (primarily power generation) and relies mainly on free quota allocation. Environmental taxes and fees suffer from low collection standards and significant regional variations. Green finance lacks unified evaluation criteria and innovative financing instruments, while environmental subsidies show weak correlation between incentive standards and emission reduction outcomes. These institutional deficiencies impede the transmission of market price signals, hampering effective guidance of corporate emission reduction behavior. Finally, existing research indicates that management often exhibits myopic behavior in environmental governance [19], amplifying the negative impacts of insufficient innovation incentives and market failures. Specifically, while market-based environmental regulations grant enterprises autonomy in decision-making, this flexibility paradoxically reinforces short-term orientation under imperfect institutional conditions. Faced with high innovation costs and market mechanism uncertainties, enterprises tend to opt for low-cost market transactions to avoid emission reduction responsibilities rather than investing in long-term technological innovation. This “inertia effect”, caused by the combination of innovation’s dual externalities, institutional deficiencies, and myopic behavior, leads enterprises to rely more on simple market transactions than substantive efficiency improvements during the market cultivation phase, ultimately resulting in increased rather than decreased carbon emission intensity.

The “incentive effect” involves the Porter hypothesis, which posits that appropriately designed environmental regulation can drive enterprises to improve resource allocation efficiency [20]. As market mechanisms continuously improve, relying solely on market transactions will bring continuously rising environmental costs. Existing research shows that market-based environmental regulation can guide enterprises to optimize resource allocation through price signals [21] to achieve emission reduction targets. On one hand, as carbon markets mature, price discovery functions gradually strengthen, and continuously rising carbon prices transmit clear emission reduction signals to enterprises; on the other hand, market mechanisms can stimulate positive competition among enterprises, with low-carbon enterprises able to gain additional benefits by selling surplus quotas. When market price signals are strong enough, enterprises recognize the importance of reducing long-term environmental costs through improved energy efficiency and actively optimize energy structures and production processes. The flexibility characteristic of market-based environmental regulation also allows enterprises to choose optimal emission reduction paths according to their own conditions, reducing carbon emission intensity.

Although enterprises show obvious “inertia effects” during initial market cultivation, as market mechanisms increasingly improve, “market-driven effects” gradually emerge, ultimately leading to decreased carbon emission intensity. Specifically, in the initial regulatory phase, the “inertia effect” plays a dominant role because enterprises view market mechanisms as cost transfer tools, tending to avoid emission reduction responsibilities through market transactions. At this time, market mechanisms are immature, price signals are weak, and enterprises can meet compliance requirements simply by purchasing quotas or paying fees, lacking motivation for emission reduction. However, as market transactions become increasingly active, the “incentive effect” gradually emerges. Continuously rising environmental costs make expenditures solely relying on market transactions climb steadily, while improved market mechanisms provide considerable economic incentives for enterprise emission reduction. Through early market experience accumulation, enterprises gradually recognize the long-term benefits of improving energy efficiency. When market mechanisms further mature, optimizing production efficiency becomes more economical compared to continuous environmental expenditures, and enterprises begin to actively reduce carbon emission intensity through technological transformation and management innovation. At this point, the “incentive effect” gradually strengthens, driving continuous decrease in enterprise carbon emission intensity. In summary, these two opposing effects result in an inverted U-shaped relationship between market-based environmental regulation intensity and enterprise carbon emission intensity (Figure 1). Therefore, this paper proposes the following hypothesis:

H1. 

Market-based environmental regulation and enterprise carbon emission intensity exhibit an inverted U-shaped relationship, first increasing then decreasing.

2.2. Market-Based Environmental Regulation and Enterprise Green Innovation

Market-based environmental regulation exhibits two effects on enterprise green innovation: a “wait-and-see effect” and a “market-driven effect”. The “wait-and-see effect” leads to a negative correlation between environmental regulation and enterprise green innovation, where stronger regulation results in fewer resources and energy invested in green innovation. The “market-driven effect” leads to a positive correlation, where stronger regulation results in greater attention and investment in green innovation.

Research based on the green paradox theory provides a rational explanation for the “wait-and-see effect” in market-based environmental regulations [16]. This effect manifests through multiple mechanisms: First, green innovation faces dual market failure challenges. Regarding environmental externality, the ecological benefits from innovation possess public good characteristics, making it difficult for innovators to receive adequate market compensation. In terms of knowledge externality, innovation outcomes exhibit high spillover effects, preventing enterprises from fully appropriating research and development returns [17,18]. These dual market failures are particularly pronounced during the exploratory phase of market-based environmental regulations. Due to insufficient institutional design and implementation experience, existing policy instruments struggle to effectively correct externality issues. Specifically, the carbon emissions trading system has limited coverage and primarily relies on free quota allocation, environmental taxes and fees maintain low standards with significant regional variations, green finance lacks unified evaluation criteria, and environmental subsidies show weak correlation with emission reduction outcomes. These institutional deficiencies not only increase information search costs and transaction uncertainties but, more critically, fail to provide sufficient innovation incentives to compensate for losses from dual externalities. Combined with green innovation’s inherent characteristics of high investment requirements, long development cycles, and significant risks [22], enterprises tend to observe market trends and postpone innovation decisions. Furthermore, regulatory lag in market supervision induces corporate “greenwashing” behavior. Under delayed environmental information disclosure supervision, enterprises may satisfy compliance requirements through selective disclosure of environmental information or exaggeration of governance achievements [23]. The cost of such superficial compliance is substantially lower than substantive technological innovation investment. More seriously, this “greenwashing” behavior not only exacerbates environmental externality issues but also distorts innovation resource allocation by misleading investor judgment [24].This results in capital misallocation to “greenwashing” enterprises, making it difficult for genuinely innovative enterprises to secure adequate financing support. This further reinforces the negative impact of knowledge externality, ultimately creating a vicious cycle of insufficient innovation momentum.

The “market-driven effect” is primarily based on the Porter hypothesis, which posits that appropriately designed environmental regulation can stimulate enterprise innovation motivation and bring innovation compensation [20]. As market mechanisms continuously improve, market-based environmental regulation transmits innovation pressure through price mechanisms, driving enterprises to seek innovative breakthroughs [25]. On one hand, as environmental rights trading markets mature, price discovery functions strengthen, and continuously rising environmental costs force enterprises to seek innovative solutions; on the other hand, market mechanisms can stimulate innovation competition among enterprises [26], with innovative enterprises able to achieve innovation compensation through selling surplus quotas or obtaining tax reductions. When market price signals are strong enough, enterprises recognize the importance of reducing long-term environmental costs through innovation and actively invest innovation resources. The flexibility characteristic of market-based environmental regulation also allows enterprises to choose optimal innovation paths according to their own conditions, improving innovation efficiency.

Although enterprises show obvious “wait-and-see effects” during initial market cultivation, as market mechanisms increasingly improve, the innovation compensation effect under the Porter hypothesis gradually emerges, ultimately leading to enhanced green innovation. Specifically, in the initial regulatory phase, the “wait-and-see effect” plays a dominant role because enterprises view market mechanisms as cost transfer tools, tending to avoid innovation risks through market transactions. At this time, market mechanisms are immature, price signals are weak, and enterprises can meet compliance requirements simply by purchasing quotas or paying fees, lacking innovation motivation. However, as market transactions become increasingly active, the “market-driven effect” gradually emerges. Continuously rising environmental costs make expenditures solely relying on market transactions climb steadily, while improved market mechanisms provide considerable economic compensation for enterprise innovation. Through early market experience accumulation, enterprises gradually recognize the long-term benefits of innovation. When market mechanisms further mature, innovation investment becomes more economical compared to continuous environmental expenditures, and enterprises begin to actively conduct green technology R&D to gain competitive advantages. At this point, the “market-driven effect” under the Porter hypothesis gradually strengthens, driving continuous improvement in enterprise green innovation levels. In summary, these two opposing effects result in a U-shaped relationship between market-based environmental regulation intensity and enterprise green innovation (Figure 2). Therefore, this paper proposes the following hypothesis:

H2. 

Market-based environmental regulation and enterprise green innovation exhibit a U-shaped relationship.

2.3. The Mediating Role of Enterprise Green Innovation

Corporate green innovation is a key driving force in reducing carbon emission intensity. On one hand, green innovation provides technical support and intellectual support for solving carbon emission intensity problems, which is an important means for enterprises to achieve low-carbon transformation; on the other hand, green innovation can alleviate the contradiction between economic growth and carbon emission intensity, helping enterprises achieve emission reduction goals while maintaining competitiveness. From the development perspective, enterprise green innovation improves energy efficiency through production process transformation and clean production technology development [27], while the application of new energy-saving technologies and clean energy significantly reduces enterprise energy consumption [28,29]. With the breakthrough of green innovation technology and the scaled application of innovation achievements, it not only brings differentiated competitive advantages [30,31,32] but also promotes the extension of green innovation in the industrial chain [33,34,35], promoting the low-carbon development of the entire industry.

Market-based environmental regulations influence carbon emission intensity through their effect on corporate green innovation behavior. During the initial phase of market mechanisms, given the characteristics of green innovation—high investment requirements, long development cycles, and significant risks—coupled with imperfect market mechanisms due to insufficient institutional design and implementation experience, enterprises face high information search costs and transaction uncertainties. During this period, enterprises exhibit a “wait-and-see effect”, tending to postpone innovation decisions or even engage in “greenwashing”, resulting in insufficient innovation momentum. As market mechanisms mature and price signals strengthen, the “market-driven effect” based on the Porter hypothesis gradually emerges. Rising environmental costs and substantial innovation compensation jointly motivate enterprises to increase green innovation investments, ultimately achieving enhanced innovation levels.

Simultaneously, corporate green innovation demonstrates significant negative impacts on carbon emission intensity. In the initial stage, enterprises improve energy efficiency through process optimization and equipment upgrades; in the development stage, enterprises develop low-carbon products and improve carbon management systems; in the mature stage, breakthrough technological innovations drive low-carbon transformation across the industrial chain. Innovation activities continuously reduce carbon emission intensity through technological innovation, product innovation, and management innovation pathways.

Therefore, corporate green innovation plays a crucial mediating role in how market-based environmental regulations influence carbon emission intensity. Market mechanisms affect corporate innovation willingness and behavior through price signals and economic incentives, while corporate innovation reduces carbon emission intensity through multiple pathways, ultimately achieving regulatory objectives. This transmission mechanism of “market incentives-innovation drive-emission reduction and efficiency improvement” reflects the inherent logic of how market-based tools achieve emission reduction effects. In summary, market-based environmental regulations stimulate enterprise green innovation through price mechanisms and economic incentives. This innovation, in turn, enhances energy efficiency and reduces carbon emission intensity. Therefore, enterprise green innovation is the key mechanism through which market-based environmental regulation impacts carbon emissions. Based on this, the hypothesis is proposed:

H3. 

Enterprise green innovation plays a mediating role between market-based environmental regulation and carbon emission intensity.

3. Research Design

3.1. Sample and Data Sources

This study covers the period from 2011 to 2020. The year 2011 marks the first year of China’s “12th Five-Year Plan” implementation, when carbon emission intensity reduction was formally incorporated as a binding target, signaling the systematic advancement of market-based environmental regulations. Selecting 2011 as the starting point enables comprehensive observation of market-based policy instruments’ evolution from inception to gradual improvement. Furthermore, 2020 holds special significance for environmental regulation research. In this year, China made its “dual carbon” commitment to the international community, marking a new development phase for market-based environmental regulations. Additionally, 2020 concluded the “13th Five-Year Plan”, when policy implementation effects had become evident and data were relatively complete.

The research sample is based on 293 prefecture-level cities published in the 2020 “Urban Statistical Yearbook”. To ensure sample consistency, cities that were elevated to prefecture-level status after 2011 were excluded. Additionally, samples with severely missing data, such as Lhasa and Sansha, were eliminated, resulting in a final analysis sample of 278 prefecture-level cities. Data on market-based environmental regulations primarily come from officially published government documents and reports, while green innovation data are sourced from the National Intellectual Property Administration. Other data are mainly derived from the “China Urban Statistical Yearbook”, prefecture-level city statistical yearbooks, and prefecture-level city national economic and social development statistical bulletins.

3.2. Variable Selection and Measurement

(1)

Carbon Emission Intensity (CI)

Following existing research [36,37,38], this study measures carbon emission intensity as the ratio of carbon emissions to GDP. The carbon emission intensity calculation formula is as follows:

$$CI={CO}_2/GDP$$

(1)

where ${CO}_2$ calculations follow IPCC guidelines. According to the IPCC (2006), this study selects coal, fuel oil, crude oil, coke, kerosene, gasoline, diesel, and natural gas as carbon emission sources. The specific measurement method is as follows:

$$C{O}_2={\displaystyle \sum _{i=1}^{n}C{O}_{2,i}}={\displaystyle \sum _{i=1}^n{E}_i}\times NC{V}_i\times CE{F}_i\times CO{F}_i\times {\displaystyle \frac{44}{12}}$$

(2)

where ${CO}_2$ represents the amount of carbon emissions (unit: 10,000 tons),${CO}_{2,i}$ is the carbon emissions of i energy, $E_i$ is the energy consumption (unit:10,000 tons), ${NCV}_i$ is the average low-calorie value (kJ/kg), ${CEF}_i$ is the carbon emission factor, and ${COF}_i$ is the carbon oxidation factor. Then, the value of CP (billion yuan/10,000 tons) is obtained.

(2)

Market-Based Environmental Regulation (market)

This study employs policy text analysis to measure the intensity of market-based environmental regulations. The specific measurement steps are as follows:

Step One: Policy Text Collection. The study period spans 2011–2020, a crucial phase in China’s advancement of ecological civilization construction and development of modern environmental governance systems, accumulating rich experience in carbon emission control policy implementation. For high-frequency word selection, we first extracted an initial keyword library from 200 environmental policy documents, then utilized NLTK to conduct word segmentation and frequency statistics on policy texts from 2011–2020, and finally formed a final set of 26 high-frequency words incorporating expert opinions. These words include “low carbon, carbon peak, carbon neutrality, carbon emissions, carbon trading” and others, with the complete word list provided in the Supplementary Materials. The search ultimately yielded 4465 policy texts encompassing types such as “opinions”, “notices”, “decisions”, and “plans”.

Step Two: Environmental Policy Screening. Based on the theoretical foundation of market-based environmental regulation, which refers to policy instruments that utilize market mechanisms to regulate corporate environmental behavior, this study focused on policy tools including carbon emission fees, carbon emission trading rights, low-carbon product subsidies, carbon taxation, and carbon finance mechanisms.

Step Three: Content Analysis Unit Determination. Taking into account both research objectives and operational feasibility, the analysis unit was defined as specific regulatory provisions within policy texts. Through systematic review of 4465 policy documents, a three-tier coding system of “document number-chapter-specific provision” was established based on the frequency and relevance of regulatory tools.

Step Four: Policy Attribute Intensity Assessment. Based on the legal effectiveness of policy texts, combined with the administrative level of policy-making institutions and text types, a five-level scoring system was established: local regulations (ordinances, provisions) 5 points; government rules (regulations, measures, detailed rules) 4 points; government administrative documents (programs, plans, methods) 3 points; government guidance documents (opinions, notices) 2 points; departmental rules (opinions, notices) 1 point.

Step Five: Policy Implementation Intensity Assessment. A quantitative standard was constructed encompassing dimensions such as action plan support, indicator constraints, and responsibility assessment specificity to evaluate the implementation intensity of carbon reduction policies.

Step Six: Data Processing. This was conducted in three steps. First, policy formulation strength of market-based environmental regulation was calculated for each environmental policy text based on policy effectiveness. Second, implementation supervision strength of market-based environmental regulation was calculated for all prefecture-level cities over the years based on evaluation standards related to environmental target responsibility and assessment systems. Third, after obtaining these two data sets, Formula (3) was used to calculate annual values of market-based environmental regulation intensity for 278 prefecture-level cities from 2011 to 2020, generating panel data of market-based environmental regulation intensity for each government-level city over the years.

$${TEP}_{ij}=\sum {PEA}_{ij}\ast P_{ij} i\in [2011, 2020]$$

(3)

where i represents the year, N represents the number of policies issued in year i, and j represents the jth policy issued in year i. ${\mathit{P}}_{\mathit{i}\mathit{j}}$ represents the policy effectiveness of the jth provision. ${\mathit{P}\mathit{E}\mathit{A}}_{\mathit{i}\mathit{j}}$ represents the policy strength of environmental target responsibility and assessment evaluation system in year i. ${\mathit{T}\mathit{E}\mathit{P}}_{\mathit{i}\mathit{j}}$ can then represent market-based environmental policy intensity in year i. In practice, as long as an environmental policy is not abolished by the government, it continues to affect real carbon dioxide emissions. Therefore, the effectiveness of environmental policies in reality is not just from environmental policies issued in the current year but accumulated from all effective environmental policies up to a certain point. Thus, when measuring, the stock of environmental policies must also be considered, with appropriate adjustments made based on policy validity periods, modifications, and abolitions (see Supplementary Materials for detailed steps).

(3)

Mediating Variable

Enterprise Green Innovation (green): This variable measures enterprise innovation activities in green technology, quantified through “number of green patent applications”.

(4)

Control Variables

Economic Growth (growth): Based on the Environmental Kuznets Curve theory, the relationship between economic development level and environmental pollution follows an inverted U-shaped pattern [39]. This study uses GDP growth rate rather than its quadratic term as a control variable, primarily because while the Environmental Kuznets Curve focuses on the inverted U-shaped relationship between economic development level and environmental pollution using quadratic terms to capture turning points, GDP growth rate as a flow indicator is only used to control for the impact of regional economic growth rate differences on carbon emission intensity. As there is no need to verify nonlinear relationships, using a linear term both satisfies control purposes and avoids model complexity and multicollinearity issues. Accordingly, this paper controls for the impact of different regional economic growth levels on carbon emission intensity through “regional GDP growth rate”.

Population Density (pop): Previous research indicates that human production activities have detrimental effects on the environment [40]. Rapid population growth can lead to poor governance, internal conflicts, or policy distortions, subsequently affecting the local natural environment [41]. Based on this, this paper uses “the logarithm of total population per unit area” to measure population density to control for the impact of population size on carbon emissions.

Foreign Direct Investment (fdi): Some scholars, drawing from pollution haven and pollution paradise hypotheses, suggest that developing countries in their early stages often attract foreign investment through lower environmental standards. However, this investment pattern may result in developed countries relocating heavily polluting industries to developing countries, ultimately turning developing countries into pollution havens for developed nations [42]. Therefore, the scale of foreign investment significantly impacts environmental governance, which this paper measures using “the logarithm of actual utilized foreign direct investment”.

Human Capital (hum): Previous research indicates that human capital accumulation can simultaneously improve capital and labor productivity efficiency, thereby promoting technological progress, while improved workforce quality can enhance individual environmental awareness and subsequently improve environmental performance [43]. Accordingly, this paper measures human capital level through “the ratio of higher education students to total regional population”.

Industrial Structure (str): Existing research indicates that different industrial structures significantly impact environmental governance, primarily because secondary industry, being pollution-intensive, increases energy consumption and pollutant emissions during development, adversely affecting environmental quality, while tertiary industry primarily comprises less polluting sectors [44,45]. Based on this, this paper represents the impact of industrial structure on carbon emission intensity using “the ratio of secondary industry added value to GDP”.

In the indicator calculation process, GDP, foreign direct investment, and other indicators were adjusted using 2000 as the base year, with foreign direct investment converted to RMB using exchange rates for the respective years (

Table 1. Measurements of variables.

Variable	Acronym	Variable Measurement
Carbon Emission Intensity	CI	Total carbon emissions/GDP
Market-Based Environmental Regulation	market	Text-based measurement from policy documents
Enterprise Green Innovation	green	Number of green patent applications
Regional Economic Growth	growth	Regional GDP growth rate
Foreign Direct Investment	fdi	Logarithm of actual utilized foreign direct investment amount
Population Density	pop	Logarithm of total population per unit area
Human Capital	hum	Ratio of higher education students to total regional population
Industrial Structure	str	Ratio of secondary industry value-added to GDP

).

3.3. Model Construction

To test the impact of market-based environmental regulation on carbon emission intensity and the mediating role of enterprise green innovation, following existing research, the specific steps are as follows:

Step 1: Use Formula (4) to estimate the impact of market-based environmental regulation on carbon emissions. Here is the explained variable, ${\mathbf{C}\mathbf{I}}_{\mathit{i},\mathit{t}}$, representing carbon emission intensity; ${\mathbf{market}}_{\mathit{i},\mathit{t}}, {\mathit{m}\mathit{a}\mathit{r}\mathit{k}\mathit{e}\mathit{t}}_{\mathit{i},\mathit{t}}^2$ are core explanatory variables, representing market-based environmental regulation and its quadratic term, respectively. ${\mathit{X}}_{\mathit{i}\mathit{t}}$ represents a series of control variables. ${\mathit{\mu }}_{\mathit{i}}$ is the city fixed effect. ${\mathit{\gamma }}_{\mathit{t}}$ is the time fixed effect.

$${\mathbf{C}\mathbf{I}}_{\mathit{i},\mathit{t}}={\mathit{\alpha }}_0+{\mathit{\alpha }}_1{\mathbf{m}\mathbf{a}\mathbf{r}\mathbf{k}\mathbf{e}\mathbf{t}}_{\mathit{i},\mathit{t}}+{\mathit{\alpha }}_2{\mathit{m}\mathit{a}\mathit{r}\mathit{k}\mathit{e}\mathit{t}}_{\mathit{i},\mathit{t}}^2+{{\mathit{\lambda }}_1\mathit{X}}_{\mathit{i}\mathit{t}}+{\mathit{\mu }}_{\mathit{i}}+{\mathit{\gamma }}_{\mathit{t}}+{\mathit{\epsilon }}_{\mathit{i},\mathit{t}}$$

(4)

Step 2: Use Formula (5) to estimate the impact of market-based environmental regulation on enterprise green innovation. Here is the explained variable, ${\mathbf{g}\mathbf{r}\mathbf{e}\mathbf{e}\mathbf{n}}_{\mathit{i},\mathit{t}}$, representing enterprise green technology innovation behavior; ${\mathbf{m}\mathbf{a}\mathbf{r}\mathbf{k}\mathbf{e}\mathbf{t}}_{\mathit{i},\mathit{t}}, {\mathit{m}\mathit{a}\mathit{r}\mathit{k}\mathit{e}\mathit{t}}_{\mathit{i},\mathit{t}}^2$ are core explanatory variables, representing market-based environmental regulation and its quadratic term, respectively. Control variables ${\mathit{X}}_{\mathit{i}\mathit{t}}$ are the same as in Formula (4). ${\mathit{\mu }}_{\mathit{i}}$ is the city fixed effect. ${\mathit{\gamma }}_{\mathit{t}}$ is the time fixed effect.

$${\mathbf{g}\mathbf{r}\mathbf{e}\mathbf{e}\mathbf{n}}_{\mathit{i},\mathit{t}}={\mathit{\alpha }}_0+{\mathit{\alpha }}_1{\mathbf{m}\mathbf{a}\mathbf{r}\mathbf{k}\mathbf{e}\mathbf{t}}_{\mathit{i},\mathit{t}}+{\mathit{\alpha }}_2{\mathit{m}\mathit{a}\mathit{r}\mathit{k}\mathit{e}\mathit{t}}_{\mathit{i},\mathit{t}}^2+{{\mathit{\lambda }}_1\mathit{X}}_{\mathit{i}\mathit{t}}+{{\mathit{\mu }}_{\mathit{i}}+{\mathit{\gamma }}_{\mathit{t}}+\mathit{\epsilon }}_{\mathit{i},\mathit{t}}$$

(5)

Step 3: Use Formula (5) to estimate the mediating role of enterprise green innovation. Here in the above equation is the explained variable, ${\mathbf{C}\mathbf{I}}_{\mathit{i},\mathit{t}}$, representing carbon emission intensity; ${\mathbf{m}\mathbf{a}\mathbf{r}\mathbf{k}\mathbf{e}\mathbf{t}}_{\mathit{i},\mathit{t}}$, ${\mathit{m}\mathit{a}\mathit{r}\mathit{k}\mathit{e}\mathit{t}}_{\mathit{i},\mathit{t}}^2$ are core explanatory variables, representing market-based environmental regulation and its quadratic term, respectively; ${\mathbf{g}\mathbf{r}\mathbf{e}\mathbf{e}\mathbf{n}}_{\mathit{i},\mathit{t}}$ represents the mediating variable, representing enterprise green technology innovation. Control variables ${\mathit{X}}_{\mathit{i}\mathit{t}}$ are the same as in Formula (4). ${\mathit{\mu }}_{\mathit{i}}$ is the city fixed effect. ${\mathit{\gamma }}_{\mathit{t}}$ is the time fixed effect.

$${\mathbf{C}\mathbf{I}}_{\mathit{i},\mathit{t}}={\mathit{\alpha }}_0+{\mathit{\alpha }}_1{\mathbf{m}\mathbf{a}\mathbf{r}\mathbf{k}\mathbf{e}\mathbf{t}}_{\mathit{i},\mathit{t}}+{\mathit{\alpha }}_2{\mathit{m}\mathit{a}\mathit{r}\mathit{k}\mathit{e}\mathit{t}}_{\mathit{i},\mathit{t}}^2+{{\mathit{\alpha }}_3\mathbf{g}\mathbf{r}\mathbf{e}\mathbf{e}\mathbf{n}}_{\mathit{i},\mathit{t}}+{{\mathit{\lambda }}_1\mathit{X}}_{\mathit{i}\mathit{t}}+{\mathit{\mu }}_{\mathit{i}}+{\mathit{\gamma }}_{\mathit{t}}+{\mathit{\epsilon }}_{\mathit{i},\mathit{t}}$$

(6)

4. Results

4.1. Descriptive Statistics

Table 2. Descriptive statistics of the main variables.

	Count	Mean	Sd	Min	Max
CI	2780	0.028	0.027	0.002	0.153
market	2780	0.995	1.239	0.000	3.761
green	2780	4.917	1.574	1.609	8.826
pop	2780	5.721	0.911	2.864	7.200
growth	2780	10.720	0.553	9.455	12.052
fdi	2780	0.016	0.016	0.000	0.070
str	2780	0.878	0.076	0.618	0.995
hum	2780	0.019	0.025	0.001	0.120

reports descriptive statistics for the required variables in four aspects: mean, standard deviation, minimum value, and maximum value. The carbon emission intensity (CI) has a mean value of 0.028, indicating relatively low carbon emissions per unit of economic output. Its standard deviation of 0.027 suggests relatively small differences between cities or years, with a range from a minimum of 0.002 to a maximum of 0.153, highlighting significant variations in carbon emission intensity across regions or time points. The market-based environmental regulation (market) has a mean value of 0.995, which, compared to its maximum value of 3.761, indicates relatively low average intensity. Its standard deviation of 1.239 shows substantial variation in implementation across different regions or time points, ranging from 0 (no market incentives) to 3.761 (high-intensity market incentives). Enterprise green technology innovation (green) has a mean value of 4.917, indicating a relatively high average level. The standard deviation of 1.574 reflects differences in green innovation among different enterprises, with a range from 1.609 to 8.826.

As shown in Figure 3, from 2011 to 2020, the intensity of market-based environmental regulations in eastern, central, and western China showed an overall upward trend. Among them, the eastern region, with its stronger economic foundation and higher degree of marketization, showed the most prominent overall performance. Although there were significant fluctuations in 2014–2015, it reached the highest level of about 14 by 2020. The central region, despite starting from a lower level, maintained stable growth, especially after 2018, reaching about 12 by 2020. The western region, although having the lowest overall level, showed a steady upward trend, increasing from about 3 in 2011 to about 10 in 2020. Notably, after 2016, the regulatory intensity gap between the three regions gradually narrowed, especially after the announcement of the “dual carbon” goals in 2020, when the market-based environmental governance levels of all three regions improved significantly, which is both a result of the deepening reform of market-based environmental governance and a positive outcome of China’s regional coordinated development strategy.

The above figure displays the spatial distribution of carbon emission intensity at the prefecture-city level in China, generated using ArcGIS 10.8 software. Based on Figure 4, we observe significant temporal and spatial heterogeneity in carbon emission intensity patterns. From a temporal perspective, high-intensity regions are primarily concentrated in central China, with their scope gradually diminishing, reflecting carbon reduction achievements in heavy industrial areas. Low-intensity regions maintain relative stability in western and eastern coastal areas, indicating either inherently cleaner industrial structures or successful industrial transformation in these regions. Medium-intensity regions show slight expansion in eastern areas, revealing potential impacts of industrial transfer. From a spatial distribution perspective, a distinct “three-zone” pattern emerges: The eastern coastal zone predominantly features low carbon emission intensity, benefiting from advanced industrial structures and technological levels; the central transition zone shows concentrated high-intensity areas, primarily due to heavy industry concentration and coal-dominated energy structures; the vast western zone generally maintains low intensity, closely related to its relatively lower industrialization level. The evolution trend from 2011 to 2020 indicates gradually narrowing regional disparities in carbon emission intensity, with high-intensity areas becoming more spatially concentrated and low-intensity areas expanding. This transformation trend aligns with China’s strategic goals of carbon peak and carbon neutrality.

4.2. Baseline Regression

Table 3. Mediating effects of green innovation.

	(1)	(2)	(3)
	CI	green	CI
market	0.006 ***	−0.416 ***	0.002 **
	(0.001)	(0.049)	(0.001)
market2	−0.207 ***	16.070 ***	−0.079 *
	(0.043)	(1.654)	(0.042)
green			−0.008 ***
			(0.000)
pop	−0.010 ***	0.673 ***	−0.005 ***
	(0.000)	(0.019)	(0.001)
growth	−0.011 ***	0.982 ***	−0.004 ***
	(0.001)	(0.047)	(0.001)
fdi	−0.089 ***	5.292 ***	−0.047 *
	(0.029)	(1.091)	(0.027)
str	−0.046 ***	0.468	−0.042 ***
	(0.008)	(0.324)	(0.008)
hum	0.011	12.387 ***	0.110 ***
	(0.019)	(0.744)	(0.019)
_cons	0.247 ***	−10.192 ***	0.166 ***
	(0.010)	(0.392)	(0.011)
City fixed effect	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes
N	2780	2780	2780

Note: Standard errors in brackets, * p < 0.1, ** p < 0.05, *** p < 0.01.

provides detailed reporting of the impact effects of market-based environmental regulation on carbon emission intensity, providing a basis for in-depth analysis. Through careful interpretation of the data in

Table 4. Analysis excluding 2020 data.

	(1)	(2)	(3)
	CI	green	CI
market	0.006 ***	−0.435 ***	0.002 *
	(0.001)	(0.052)	(0.001)
market2	−0.205 ***	16.518 ***	−0.075 *
	(0.046)	(1.761)	(0.044)
green			−0.008 ***
			(0.000)
pop	−0.010 ***	0.689 ***	−0.005 ***
	(0.001)	(0.020)	(0.001)
growth	−0.011 ***	0.980 ***	−0.004 ***
	(0.001)	(0.050)	(0.001)
fdi	−0.091 ***	5.662 ***	−0.047
	(0.031)	(1.188)	(0.030)
str	−0.045 ***	0.376	−0.042 ***
	(0.009)	(0.347)	(0.009)
hum	0.013	12.222 ***	0.108 ***
	(0.021)	(0.798)	(0.021)
_cons	0.246 ***	−10.255 ***	0.166 ***
	(0.011)	(0.413)	(0.011)
City fixed effect	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes
N	2502	2502	2502

Note: Standard errors in brackets, * p < 0.1, ** p < 0.05, *** p < 0.01.

, the following analysis can be derived:

The data in column (1) of Table 3 indicate that the impact of market-based environmental regulations on carbon emission intensity exhibits a similar inverted “U-shaped” nonlinear characteristic. The regression coefficient of the first-order term is 0.006, while the second-order term coefficient is −0.207, both significant at the 1% level. This suggests that during the initial stage of market-based environmental regulations, carbon emission intensity might slightly increase due to increased economic activity, but as environmental regulation policies are implemented more deeply, carbon emission intensity gradually decreases. This result validates Hypothesis H1, demonstrating that market-based environmental regulations can effectively regulate carbon emission intensity.

The data in column (2) of Table 3 show that the first-order and second-order term coefficients of market-based environmental regulations are −0.416 and 16.070, respectively, passing significance level tests to varying degrees. This result indicates that market-based environmental regulations have a “U-shaped” impact on enterprise green innovation. Hypothesis H2 is thus validated.

The data in column (3) of Table 3 reveal that the regression coefficient of enterprise green innovation on carbon emission intensity is significantly negative. Meanwhile, the first-order and second-order term coefficients of market-based environmental regulations are 0.002 and −0.079, respectively, passing significance level tests. This result indicates that market-based environmental regulations also exert their inverted “U-shaped” influence on carbon emission intensity by promoting enterprise green innovation.

4.3. Robustness Tests

(1)

Controlling for Pandemic Impact

Given the significant impact of the global COVID-19 pandemic in 2020 on economic and social activities, which may have led to abnormal data fluctuations, this paper conducted robustness tests excluding 2020 data to analyze more stable economic and social patterns in other years. By excluding data from this exceptional year, this paper re-evaluated the impact of market-based environmental regulation on carbon emission intensity and the mediating role of enterprise green innovation. As shown in Table 4, the regression results are consistent with the previous findings, confirming the robustness of our analysis.

(2)

Excluding Sub-provincial Cities

Considering that sub-provincial cities have certain advantages over other types of cities in terms of administrative level, economic development level, concentration of innovation factors, and innovation capabilities, these differences might lead to some bias when analyzing overall data, thereby affecting the accuracy and general applicability of this paper’s conclusions. After excluding data from sub-provincial cities, this paper conducted regression analysis again, and according to

Table 5. Analysis excluding sub-provincial cities.

	(1)	(2)	(3)
	CI	green	CI
market	0.006 ***	−0.404 ***	0.003 **
	(0.001)	(0.049)	(0.001)
market2	−0.215 ***	15.741 ***	−0.086 **
	(0.044)	(1.672)	(0.043)
green			−0.008 ***
			(0.000)
pop	−0.011 ***	−0.010 ***	−0.005 ***
	(0.001)	(0.001)	(0.001)
growth	−0.013 ***	−0.012 ***	−0.004 ***
	(0.001)	(0.001)	(0.001)
fdi	−0.099 ***	−0.078 ***	−0.032
	(0.029)	(0.029)	(0.028)
str	−0.042 ***	−0.045 ***	−0.040 ***
	(0.009)	(0.009)	(0.008)
hum	0.001	0.007	0.111 ***
	(0.020)	(0.020)	(0.020)
_cons	0.257 ***	0.254 ***	0.173 ***
	(0.010)	(0.011)	(0.011)
City fixed effect	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes
N	2660	2660	2660

Note: Standard errors in brackets, * p < 0.1, ** p < 0.05, *** p < 0.01.

, the results remained consistent with the previous analysis.

(3)

Replacing the Explained Variable

To further ensure the robustness and reliability of the research results, this paper employed per capita carbon emission intensity as an alternative indicator to measure carbon emission intensity. According to

Table 6. Alternative explained variable analysis.

	(1)	(2)	(3)
	CI	green	CI
market	0.001 ***	−0.416 ***	0.001 **
	(0.000)	(0.049)	(0.000)
market2	−0.039 ***	16.070 ***	−0.018 *
	(0.010)	(1.654)	(0.010)
green			−0.001 ***
			(0.000)
pop	−0.003 ***	0.673 ***	−0.002 ***
	(0.000)	(0.019)	(0.000)
growth	−0.002 ***	0.982 ***	−0.000
	(0.000)	(0.047)	(0.000)
fdi	−0.017 **	5.292 ***	−0.010
	(0.007)	(1.091)	(0.006)
str	−0.014 ***	0.468	−0.013 ***
	(0.002)	(0.324)	(0.002)
hum	0.007	12.387 ***	0.023 ***
	(0.005)	(0.744)	(0.005)
_cons	0.053 ***	−10.192 ***	0.040 ***
	(0.002)	(0.392)	(0.003)
City fixed effect	Yes	Yes	Yes
Year fixed effect	Yes	Yes	Yes
N	2780	2780	2780

Note: Standard errors in brackets, * p < 0.1, ** p < 0.05, *** p < 0.01.

, the results obtained remain consistent with the previous analysis.

(4)

Construct interaction terms to verify mediation effects

This paper constructs interaction terms between mediating variables and explanatory variables, incorporating them into the model to verify whether the mediating effect remains significant. Column (2) of

Table 7. Alternative construct interaction terms to verify mediation effects.

	(1)	(2)
	CI	CI
green	−0.008 ***	−0.008 ***
	(0.000)	(0.000)
market		0.003 **
		(0.001)
market2		−0.114 **
		(0.044)
market_green		−0.010 ***
		(0.002)
market2_green		0.010 ***
		(0.001)
pop	−0.005 ***	−0.005 ***
	(0.001)	(0.001)
growth	−0.004 ***	−0.004 ***
	(0.001)	(0.001)
fdi	−0.048 *	−0.051 *
	(0.027)	(0.027)
str	−0.041 ***	−0.042 ***
	(0.008)	(0.008)
hum	0.107 ***	0.095 ***
	(0.019)	(0.019)
_cons	0.168 ***	0.168 ***
City fixed effect Year fixed effect	(0.011) Yes Yes	(0.011) Yes Yes
N	2780	2780

Note: Standard errors in brackets, * p < 0.1, ** p < 0.05, *** p < 0.01.

reveals how market-based environmental regulations similarly influence carbon emission intensity through the mediating variable of enterprise green innovation. Likewise, the regression coefficient of enterprise green innovation on carbon emission intensity is significantly negative. The first-order and second-order term coefficients of market-based environmental regulations are 0.002 and −0.079, respectively, while their interaction term coefficients with enterprise green innovation are −0.010 and 0.010, respectively, all passing significance level tests to varying degrees. This result indicates that market-based environmental regulations also exert their inverted “U-shaped” influence on carbon emission intensity by promoting enterprise green innovation.

4.4. Endogeneity Analysis

Although this study controlled for variables that might affect empirical results during model construction and enhanced research conclusion reliability through a series of robustness tests, concerns remain about potential reverse causality between environmental regulation intensity and carbon emission intensity, potentially undermining research conclusion credibility. In addressing the endogeneity issue between market-based environmental regulation and carbon emission intensity, enterprise registration numbers were chosen as instrumental variables for the following reasons: The first is high correlation. Enterprise registration numbers are an important indicator measuring a city’s market development level and economic activity vitality. Higher enterprise registration numbers indicate more developed market systems and active market entities, providing a solid micro-foundation for implementing market-based environmental regulation policies. Specifically, increased enterprise numbers help form more sufficient emissions trading markets and enhance environmental rights liquidity; meanwhile, they create better market conditions for implementing market-based policy tools like green credit and environmental taxes. Additionally, increased enterprise registration numbers reflect local governments’ emphasis on market-oriented reform, demonstrating stronger market-oriented governance orientation in carbon emission management, helping enhance market-based environmental regulation policy intensity. The second is strict exogeneity. Enterprise registration numbers are mainly influenced by factors such as regional overall business environment, market access policies, and administrative approval efficiency, which have no direct causal relationship with carbon emission intensity. Furthermore, carbon emission intensity, as carbon emissions per unit GDP, depends more on factors like industrial structure and technological level rather than enterprise numbers. Moreover, enterprise registration numbers have certain historical evolution and institutional path dependency, not significantly affected by current carbon emission intensity, further ensuring their exogeneity.

This paper introduces instrumental variables to avoid direct endogenous associations between environmental regulation and carbon emission intensity. The selection of instrumental variables is based on their correlation with the explanatory variable (i.e., market-based environmental regulation) while maintaining no direct association with the error term, thereby effectively avoiding endogeneity issues. Specifically, this paper uses the number of registered enterprises in cities as an instrumental variable for market-based environmental regulation; the number of registered enterprises is an important indicator measuring a city’s economic activity vitality and market development level. Market-based environmental regulation typically influences enterprise pollution behavior through market mechanisms (such as emissions trading rights and green credit). The increase in enterprise numbers promotes market improvement, thereby facilitating effective implementation of market-based environmental regulation. Therefore, there exists an indirect but significant correlation between enterprise registration numbers and market-based environmental regulation. Although enterprise registration numbers may be influenced by multiple factors, they more reflect a city’s overall economic environment and market conditions rather than directly affecting carbon emission intensity relative to specific city carbon emission intensity and thus can be considered an exogenous variable.

To ensure instrumental variable validity, this paper employed the Anderson canonical correlations LM statistic for endogeneity testing. The core of this testing method lies in determining whether instrumental variables correlate with explanatory variables while remaining uncorrelated with error terms. A significant LM statistic indicates that the selected instrumental variables are valid and successfully address endogeneity issues. To further verify instrumental variable validity, this paper also used the Cragg–Donald Wald F statistic for weak instrumental variable testing. This test aims to determine whether the selected instrumental variables are sufficiently “strong”, i.e., whether they adequately correlate with explanatory variables. A significant F statistic at the set significance level indicates sufficiently strong instrumental variables, ensuring IV (instrumental variable) estimation consistency and effectiveness. In this paper’s research, F statistics are significant at the 5% level, further confirming the validity of the instrumental variables.

Through adopting the above methods and strategies, this paper successfully addressed the endogeneity issue between market-based environmental regulation and carbon emission intensity, obtaining more accurate and reliable model estimation results. Not only are these results statistically significant, their regression directions also remain consistent with previous analyses, further proving the robustness and reliability of this paper’s research hypotheses and conclusions (

Table 8. Endogeneity test results.

	CI
market	0.177 ***
	(0.019)
market2	−5.797 ***
	(0.617)
pop	−0.005 ***
	(0.001)
growth	0.012 ***
	(0.004)
fdi	0.048
	(0.079)
str	−0.111 ***
	(0.024)
hum	0.328 ***
	(0.063)
_cons	0.004
	(0.038)
idstat	94.896
widstat	97.649
N	2780.000

Note: Standard errors in brackets, * p < 0.1, ** p < 0.05, *** p < 0.01.

).

5. Analysis of Official Heterogeneity

After verifying the impact and mechanisms of market-based environmental regulations on carbon emission intensity, an important question emerges: do these effects vary significantly across different regions? Given the characteristics of China’s distinctive cadre management system, local officials, as key actors in policy formulation and implementation, may significantly influence the effectiveness of market-based environmental regulations through their personal characteristics [46]. For instance, officials of different age groups may have varying levels of understanding and capability in utilizing market-based instruments, while their educational background might affect their receptiveness to innovation policies. Additionally, local and non-local officials face different information constraints and social capital support during policy implementation, which may cause identical market-based environmental regulations to yield different effects across regions. Therefore, an in-depth analysis of the heterogeneous impacts of official characteristics not only helps us better understand regional differences in policy effectiveness but also provides important insights for improving cadre management systems and enhancing policy implementation outcomes. Consequently, to examine whether individual characteristics of different officials influence the relationship between market-based environmental regulations and carbon emissions, we selected four official characteristics for further analysis: tenure, education level, age, and origin. Data on official characteristics were primarily sourced from public online materials, including Xinhua Net, People’s Daily Online, and Baidu Encyclopedia, with contemporary news reports serving as the primary reference when the exact starting year of mayoral appointments was not explicitly stated.

5.1. Official Age

Existing research indicates that once prefectural-level city secretaries and mayors exceed 54 years of age, their probability of promotion decreases significantly, while the likelihood of stepping down to secondary positions increases substantially [47]. To examine the influence of officials of different ages, this paper groups government-level city mayors by age, with the dummy variable age assigned a value of 1 if the official is under 54 years old and 0 otherwise, to separately test the impact of market-based environmental regulation on carbon emission intensity.

As shown in columns (1) and (2) of

Table 9. Heterogeneity tests based on official age.

	(1)	(2)
	Age = 1	Age = 0
market	0.007 ***	0.002
	(0.002)	(0.002)
market2	−0.266 ***	−0.076
	(0.053)	(0.078)
pop	−0.011 ***	−0.007 ***
	(0.001)	(0.001)
growth	−0.010 ***	−0.015 ***
	(0.001)	(0.002)
fdi	−0.103 ***	−0.029
	(0.034)	(0.052)
str	−0.033 ***	−0.101 ***
	(0.010)	(0.017)
hum	−0.014	0.075 **
	(0.025)	(0.031)
_cons	0.229 ***	0.321 ***
	(0.012)	(0.019)
City fixed effect	Yes	Yes
Year fixed effect	Yes	Yes
N	2160	620

Note: Standard errors in brackets, * p < 0.1, ** p < 0.05, *** p < 0.01.

, when officials are younger than 54 years old, the impact of market-based environmental regulation policies on carbon emission intensity follows a pattern of initial increase followed by suppression. This may be attributed to younger officials typically possessing stronger innovative mindsets and market-oriented concepts, making them more inclined to actively employ market-based tools such as emission trading rights, environmental taxes and fees, and green credit to promote environmental governance. During the initial implementation phase, as enterprises need to invest substantial resources to adapt to market-based environmental regulation requirements, upgrading environmental protection equipment and modifying production processes, carbon emission intensity may increase in the short term. However, as young officials continue to promote market-based environmental regulation policies, enterprises gradually adapt to policy requirements under market mechanism guidance, optimizing production methods through technological innovation and improving resource allocation efficiency, ultimately achieving sustained reduction in carbon emission intensity.

However, when officials are over 54 years old, market-based environmental regulation policies fail to impact carbon emission intensity significantly. This may be attributed to several factors: First, due to their developmental environment and work experience, they may be more accustomed to using traditional administrative measures for environmental governance, with relatively lower acceptance and operational capability regarding market-based policy instruments. Additionally, older officials tend to rely more on existing management experience during policy implementation, potentially having limited understanding and grasp of new market-based tools, making it difficult to fully leverage the resource allocation functions of mechanisms like emission trading and green finance, affecting policy guidance on enterprise behavior. Furthermore, older officials often demonstrate stronger risk-averse tendencies. As innovative governance tools, market-based environmental regulation policies inevitably involve certain uncertainties in their implementation. Older officials may prefer traditional regulatory measures with lower risk, maintaining a cautious attitude toward the application of market-based tools like emission trading and green credit, weakening the market mechanism’s role in promoting enterprise technological innovation and diminishing the carbon reduction effects of market-based environmental regulations.

5.2. Official Origin

Since the 1990s, the central government has established a series of cadre exchange systems, which have had significant impacts on talent team building and urban development. Local officials from different origins exhibit different personal experiences, which may influence the relationship between command-and-control environmental regulation and carbon emission intensity. Since the 1990s, China’s central government has established a series of official exchange systems for cadres, significantly impacting talent development and urban development.

Valuable research has found that governor exchanges increased economic growth rates in destination regions [48] and significantly reduced corruption in these areas. This is because exchanged officials need to strive for outstanding achievements to gain greater promotion opportunities. However, opposing views suggest that official exchanges may be detrimental to local development [49]. More importantly, local officials from different origins have different personal experiences, leading to varying economic governance performance, which is also influenced by age and tenure [50,51]. For example, central government officials tend to focus on investigation, training, and development goals, showing lower motivation for excellent performance during their tenure [48]. This may be because they maintain good connections with the central government and place less emphasis on local economic development. Conversely, locally promoted officials have more abundant local information, benefiting local development. Furthermore, officials promoted or transferred from other regions may strive to demonstrate their uniqueness under the dual pressure of incomplete local information and official systems [52]. To further verify the influence of official origin, this paper conducts group tests based on whether officials are local, with the dummy variable source taking a value of 1 if the official is local and 0 otherwise.

According to columns (1) and (2) of

Table 10. Heterogeneity tests based on official origin.

	(1)	(2)
	Source = 1	Source = 0
market	0.007 ***	−0.001
	(0.002)	(0.000)
market2	−0.231 ***	0.014
	(0.069)	(0.014)
pop	−0.012 ***	−0.002 *
	(0.001)	(0.001)
growth	−0.015 ***	−0.021 **
	(0.002)	(0.001)
fdi	−0.116 **	−0.024 *
	(0.045)	(0.014)
str	0.038 **	−0.043 ***
	(0.016)	(0.008)
hum	−0.043	0.022
	(0.032)	(0.028)
_cons	0.225 ***	0.3.02 ***
	(0.017)	(0.011)
City fixed effect	Yes	Yes
Year fixed effect	Yes	Yes
N	1177	1603

Note: Standard errors in brackets, * p < 0.1, ** p < 0.05, *** p < 0.01.

, when officials are of local origin, the impact of market-based environmental regulation policies on carbon emission intensity demonstrates a pattern of initial increase followed by suppression. This may be attributed to local officials having in-depth understanding of their jurisdiction’s economic and social development conditions, industrial structure characteristics, and environmental carrying capacity, enabling them to accurately gauge the regulatory intensity of market-based environmental regulation policies and reasonably utilize market-based tools such as emission trading, environmental taxes and fees, and green credit. During the initial implementation phase, as enterprises need to invest substantial resources in environmental protection equipment upgrades and production process modifications, carbon emission intensity may increase in the short term. However, as local officials continue to promote market-based environmental regulation policies, enterprises gradually adapt to policy requirements under market mechanism guidance, optimizing production methods through technological innovation and improving resource allocation efficiency, ultimately achieving sustained reduction in carbon emission intensity.

However, when officials are appointed from other regions, market-based environmental regulation policies fail to impact carbon emission intensity significantly. This may be attributed to several factors: Non-local officials lack local environmental governance experience, making it difficult to accurately grasp the application characteristics of local market-based environmental regulation tools. Furthermore, non-local officials have insufficient understanding of key challenges in local carbon emission management and require longer periods to adapt to and understand local carbon management policy characteristics, further affecting the effective implementation of market incentive policies. Additionally, non-local officials lack political connections with local enterprises, making it difficult to effectively coordinate and integrate market stakeholders’ interests, potentially leading to significant resistance in implementing market-based policies such as environmental rights trading and green finance. Simultaneously, the lack of effective communication channels with local enterprises, financial institutions, and industry associations makes it difficult to accurately understand market entities’ environmental governance needs and innovation capabilities, affecting the targeting and operability of market incentive policies in carbon emission management. Most importantly, due to insufficient local social capital support, they may face higher implementation costs and governance resistance in promoting market-based environmental regulation policies. These factors collectively make it difficult for market-based environmental regulation policies to form effective market constraints on enterprises, thereby limiting their effectiveness in carbon emission reduction processes.

5.3. Official Education Level

Education level typically refers to an individual’s degree of formal education, reflecting personal psychological qualities in work and knowledge accumulation in problem-solving [53]. Although officials at all levels in China have seen significant improvements in educational attainment since the reform and opening-up, considerable differences in education levels still exist among current officials. To further analyze the impact of officials’ education levels on the relationship between command-and-control environmental regulation and carbon emission intensity, this paper conducts group tests using a master’s degree as the educational threshold, with the dummy variable education taking a value of 1 if the official has a master’s degree or above and 0 otherwise.

According to columns (1) and (2) of

Table 11. Heterogeneity tests based on official education level.

	(1)	(2)
	Education = 1	Education = 0
market	0.006 ***	0.004 *
	(0.001)	(0.002)
market2	−0.207 ***	−0.180 **
	(0.049)	(0.083)
pop	−0.011 ***	−0.005 ***
	(0.001)	(0.001)
growth	−0.010 ***	−0.017 ***
	(0.001)	(0.002)
fdi	−0.111 ***	−0.006
	(0.032)	(0.061)
str	−0.051 ***	−0.004
	(0.010)	(0.016)
hum	0.014	−0.012
	(0.022)	(0.038)
_cons	0.246 ***	0.241 ***
	(0.011)	(0.020)
City fixed effect	Yes	Yes
Year fixed effect	Yes	Yes
N	2351	429

Note: Standard errors in brackets, * p < 0.1, ** p < 0.05, *** p < 0.01.

, officials’ education levels do not influence the relationship between market-based environmental regulations and carbon emission intensity, meaning that regardless of whether officials have high or low education levels, the impact of market-based environmental regulations on carbon emission intensity maintains an inverted U-shape. This may be attributed to several factors: First, although higher education levels might indicate more systematic knowledge reserves, in environmental governance practice, officials typically rely on professional environmental management departments, research institutions, and market intermediary organizations. These institutions possess professional advantages in emission trading system design, environmental tax standard setting, and green finance policy innovation, providing comprehensive decision-making support for officials with different educational backgrounds. Additionally, during policy implementation, officials can gain deep understanding of policy operational characteristics through participating in the specific implementation of market-based tools such as emission trading, environmental taxes and fees, and green credit. For instance, they grasp market entities’ environmental governance needs through enterprise interactions, obtain implementation effectiveness feedback through policy evaluation, and learn advanced experiences through cross-regional exchanges. This practice-based experience accumulation enables officials with different educational backgrounds to effectively manage the regulatory intensity of market-based environmental regulation policies, thereby weakening the impact of educational background differences.

5.4. Official Tenure

Tenure generally refers to the time period during which officials exercise formal power [54]. In the context of “competition for growth”, some scholars argue that newly appointed officials tend to adopt strategies that stimulate rapid economic growth in the short term, potentially adversely affecting the environment [55]. However, some research has found that longer tenures of local officials make it easier for them to establish political connections with local enterprises and protect polluting behaviors for economic growth and stable tax sources [55,56], thus hindering environmental governance efficiency improvement. A third perspective suggests that there exists an inverted U-shaped relationship between official tenure and environmental governance, with officials’ emphasis on environmental governance showing a trend of initial decrease, followed by increase and then decrease during their tenure. Cao et al. found a U-shaped relationship between party secretary tenure and air PM2.5 content [57]; Yu et al.’s research revealed a weak inverted U-shaped relationship between time constraints and economic growth, as provincial governors are more likely to take drastic measures to complete performance assessments when their terms are ending [58], sometimes at the expense of environmental protection. To further verify the influence of official tenure, this paper conducts group tests based on whether government-level city mayors’ tenures are greater than or equal to 5 years, with the dummy variable tenure = 1 if the official’s tenure is greater than or equal to 5 years and 0 otherwise.

According to columns (1) and (2) of

Table 12. Heterogeneity tests based on official tenure.

	(1)	(2)
	Tenure = 1	Tenure = 0
market	0.000	0.005 ***
	(0.001)	(0.001)
market2	−0.022	−0.201 ***
	(0.043)	(0.046)
pop	0.011	−0.010 ***
	(0.008)	(0.001)
growth	−0.013 ***	−0.011 ***
	(0.004)	(0.001)
fdi	0.099	−0.091 ***
	(0.014)	(0.030)
str	0.031 *	−0.046 ***
	(0.018)	(0.009)
hum	0.093	0.013
	(0.063)	(0.021)
_cons	0.078	0.245 ***
	(0.066)	(0.011)
City fixed effect	Yes	Yes
Year fixed effect	Yes	Yes
N	299	2481

Note: Standard errors in brackets, * p < 0.1, ** p < 0.05, *** p < 0.01.

, when officials’ tenure is less than 5 years, the impact of market-based environmental regulation policies on carbon emission intensity shows a pattern of initial increase followed by suppression. This may be attributed to officials with shorter tenures facing stronger promotion pressure and performance assessment requirements, making them more inclined to innovatively employ market-based environmental regulation tools. They actively promote market-oriented reform of environmental regulation policies through optimizing emission trading systems, improving environmental tax mechanisms, and innovating green finance policies. During initial implementation, as enterprises need to invest substantial resources to adapt to market-based environmental regulation requirements, upgrading environmental protection equipment and modifying production processes, carbon emission intensity may increase in the short term. However, as officials continue to promote market-based environmental regulation policies, enterprises gradually adapt to policy requirements under market mechanism guidance, optimizing production methods through technological innovation and improving resource allocation efficiency, ultimately achieving sustained reduction in carbon emission intensity.

However, when officials’ tenure exceeds 5 years, market-based environmental regulation policies fail to impact carbon emission intensity significantly, primarily due to several factors: First, long-term positioning in the same post may lead to path dependency, with officials developing inertial reliance on existing market-based policy tool implementation patterns, lacking flexibility in applying new market-based governance tools, and struggling to timely adjust market incentive schemes according to jurisdictional economic development levels and industrial structure characteristics. Second, officials with longer tenures may establish stable interest alliances with local key enterprises and industry associations. Some enterprises might evade environmental regulations through rent-seeking, leading to selective enforcement or weak implementation of policies. More importantly, long-established interest networks may hinder the implementation of new environmental regulation policies, preventing market incentive mechanisms from functioning effectively. Finally, from a promotion incentive perspective, longer tenure often indicates narrowed career advancement opportunities. At this point, local officials, lacking promotion pressure, may reduce their emphasis on long-term policy objectives such as environmental quality improvement, further weakening the implementation of market-based environmental regulation policies and affecting policy effectiveness.

6. Policy Recommendations

6.1. Research Conclusions

Based on policy text analysis methodology to construct a market-based environmental regulation index, this study systematically examines the impact and mechanisms of market-based environmental regulations on carbon emission intensity using panel data from Chinese prefecture-level cities from 2011–2020. The main findings are as follows:

First, there exists an inverted U-shaped relationship between market-based environmental regulations and carbon emission intensity. During the initial implementation phase, carbon emission intensity increases as enterprises face high adaptation costs and policy tools remain in an exploratory stage of institutional construction and implementation experience. Enterprises tend to maintain existing production modes through market transactions rather than implementing substantive improvements. As the institutional system gradually matures and policy implementation effectiveness improves, strengthened price signals lead enterprises to recognize the importance of reducing long-term environmental costs through improved energy efficiency, prompting them to actively optimize energy structures and production processes, ultimately achieving sustained reduction in carbon emission intensity.

Second, enterprise green innovation plays a crucial mediating role in how market-based environmental regulations affect carbon emission intensity. During the initial stage of policy tool institutional construction, enterprises demonstrate a clear “wait-and-see effect” with insufficient innovation momentum, given green innovation’s characteristics of high investment requirements, long development cycles, and significant risks, coupled with inadequate price signal transmission. As the policy tool system continuously improves, rising environmental costs and substantial innovation compensation jointly motivate enterprises to increase green innovation investment, reducing carbon emission intensity through multiple dimensions including technological innovation, product innovation, and management innovation.

Third, official characteristics significantly influence the policy effectiveness of market-based environmental regulations. Younger officials (below 54 years) and local officials better grasp the application of policy tools; officials with shorter tenures (below 5 years) more actively promote institutional innovation, achieving more significant policy effects, while officials’ education levels show no significant impact on policy effectiveness. This indicates that under China’s distinctive cadre management system, officials’ personal characteristics influence emission reduction effects of market-based environmental regulations by affecting policy implementation methods and intensity.

6.2. Research Contribution

Based on the above research findings, this study’s contributions are manifested in three main aspects:

First, this study proposes a novel systematic method for measuring market-based environmental regulations, transcending the limitations of previous research that focused only on single policy instruments. This method systematically identifies and extracts market-based policy provisions related to carbon emissions through a three-tier coding system of “document number-chapter-specific provision”; evaluates policy attribute intensity through a five-level scoring system based on legal effectiveness and administrative hierarchy; and considers policy stock in measuring policy intensity, making dynamic adjustments based on policy validity periods, modifications, and terminations, thus more accurately reflecting cumulative policy effects. The application of this measurement method reveals significant differences in emission reduction effects between single policy tools and multiple policy combinations. While existing literature generally suggests that market-based environmental regulations positively promote carbon reduction [5,6,10,11,12], when examining their comprehensive impact, this study discovers for the first time an inverted U-shaped relationship with carbon emission intensity, indicating increased carbon emission intensity during initial policy implementation. This finding suggests that the combined effects of multiple policy tools are not simply the sum of their individual effects but generate more complex dynamic impacts, providing new perspectives for deepening understanding of market-based environmental regulation mechanisms.

Second, this study reveals the transmission mechanism of market-based environmental regulations’ impact on carbon emission intensity by introducing enterprise green innovation as a mediating variable. The research finds that gradual improvement of the policy tool system can promote enterprise innovation through price signals and economic incentives, thereby achieving emission reduction targets, providing new empirical evidence for the Porter hypothesis in environmental regulation. Additionally, the observed transition from a “wait-and-see effect” to a “market-driven effect” enriches green paradox theory research.

Finally, this study incorporates official characteristics into the analytical framework, examining their influence on market-based environmental regulation policy effectiveness. The finding that officials with different characteristics show significant variations in policy tool application and implementation effectiveness extends the research boundaries of environmental regulation theory, providing new theoretical perspectives for understanding environmental governance under China’s distinctive cadre management system.

6.3. Policy Recommendations

Based on the above research findings, this study proposes the following policy recommendations to better leverage market-based environmental regulations in improving carbon emission intensity:

First, perfect the market-based environmental regulation policy tool system. The discovered inverted U-shaped relationship indicates that incomplete institutional construction may lead to policy effects contrary to expectations. Therefore, authorities should accelerate the improvement of carbon emission trading systems by expanding carbon market coverage, perfecting price formation mechanisms, and diversifying trading products; optimize environmental tax systems by setting reasonable tax rates and establishing dynamic adjustment mechanisms linked to emission reduction effects; strengthen green financial system construction by unifying credit evaluation standards and innovating financial products and services; and improve environmental subsidy policies by scientifically setting subsidy standards and increasing fund utilization efficiency. Through constructing a comprehensive policy tool system, the market mechanism’s decisive role in resource allocation can be fully realized.

Second, strengthen the support system for enterprise green innovation. Research indicates that enterprise green innovation is a crucial transmission mechanism for market-based environmental regulations to influence carbon emission intensity. It is recommended to establish special funds supporting enterprise green technology research and development, improve intellectual property protection systems, and reduce innovation risks; leverage financial institutions to innovate green financial products and broaden enterprise financing channels; and establish industry-university-research collaborative innovation platforms to promote technological achievement transformation and improve innovation efficiency. Meanwhile, it is also important to strengthen supervision of “greenwashing” behavior, improve environmental information disclosure systems, and guide innovation resources toward substantive technological innovation.

Third, optimize cadre management systems. Given that official characteristics significantly influence policy effectiveness, cadre selection and appointment mechanisms should be improved. It is recommended to set reasonable tenure limits and establish scientific cadre rotation mechanisms to avoid policy implementation inertia from excessive tenure lengths; emphasize candidates’ understanding of local economic and social development conditions during official selection and establish comprehensive policy training and experience exchange mechanisms for non-local officials; and strengthen targeted training for officials of different age groups to enhance their understanding and application capability of market-based policy tools, providing expert consultation and intellectual support. Simultaneously, evaluation systems can be improved by incorporating environmental governance objectives into assessments and establishing long-term mechanisms promoting continuous policy innovation.

6.4. Research Deficiencies and Future Prospects

First, there are certain limitations in measuring enterprise green innovation using total patent indicators at the city level. Specifically, patent output capabilities vary significantly across industries, with pharmaceutical, biotechnology, and electronics sectors typically showing higher propensity for patent applications. Although this study partially addresses this issue by including the proportion of secondary industry added value to GDP as a control variable, obtaining data on key industry shares or comparing innovation performance across similar industries in different cities would help enhance the robustness of empirical results. Future research could consider constructing more detailed city-industry-patent correspondence relationships or conducting matching analyses based on cities with similar industrial structures to achieve more accurate innovation measurements.

Second, due to space limitations, this study only examined the mediating role of enterprise green innovation in the relationship between market-based environmental regulations and carbon emission intensity, without considering green innovation levels as threshold variables to examine how different levels of green innovation affect the impact of market-based environmental policies on emission intensity. Future research will continue to explore the influence mechanisms of market-based environmental regulations on carbon emission intensity under different green innovation levels, aiming to provide more targeted suggestions for carbon emission reduction policy formulation.