The Regulatory Effect of Government Fiscal Intervention on Carbon Reduction—A System Analysis Based on Economy–Energy–Environment

Abstract

In the process of achieving carbon neutrality, the role of government fiscal intervention is becoming more prominent. This article is based on the DPSIRM theoretical model and constructs a comprehensive evaluation index system for economic growth, energy transformation, and environmental improvement, using an improved entropy method for evaluation. Subsequently, we examined the impact of the economic, energy, and environmental systems on carbon emission reduction under the threshold effect of government fiscal intervention. The results indicate that (1) the economic development and carbon emissions in Jiangsu Province show a U-shaped trend of first decreasing and then increasing, while the energy transformation shows a positive inverted U-shaped trend of first increasing and then decreasing. Environmental improvement significantly reduces carbon emissions. (2) Government fiscal intervention has a significant impact on local carbon reduction, which is known as the regulatory effect of fiscal intervention. Specifically, with the increasing degree of government fiscal intervention, energy transformation, economic growth, and environmental improvement have promoted carbon emission reduction to varying degrees, and their impact mechanisms exhibit differentiated marginal effect characteristics. Among them, the marginal effect of energy transformation on carbon reduction increases, while the marginal effect of economic growth and environmental improvement decreases.

1. Introduction

As of now, more than 150 countries and regions have proposed carbon neutrality goals, which aim to achieve carbon neutrality by the middle of the 21st century. Reducing greenhouse gas emissions is crucial for the world’s green and sustainable development [1]. Internationally, developed countries and regions such as Europe and America have also taken many actions to address climate change and promote carbon neutrality. In recent years, the United States has increased its investment and development efforts in renewable energy sources such as solar and wind power. California and other regions are actively promoting the popularization of renewable energy and implementing strict carbon emission restriction policies to facilitate energy transition [2]. The European Union has proposed a “Green Deal” aimed at achieving climate neutrality by 2050 and incorporating carbon neutrality into various areas of economic and social development [3]. China has been the world’s largest energy consumer and carbon emitter for many years [4,5]. At an important historical juncture of the “carbon neutrality” strategy, China, as a global energy consuming and carbon emitting country, shoulders greater responsibility in accelerating the layout of green and low-carbon industries. China currently employs multiple environmental policies to control air pollutants and is attempting to curb carbon emissions through various environmental regulations [6]. The government has introduced environmental policies such as a low-carbon city pilot [7]. In addition, China is placing greater emphasis on environmental pollution control, with investments in this area steadily rising. In 2019, such investments amounted to CNY 915.19 billion, constituting 0.9% of the country’s GDP. However, the emission reduction effect is still not satisfactory [8]. Effective government policies are crucial for balancing carbon emission reduction with sustained economic growth and improved living standards [9]. The government plays an important role in economic growth, energy transition, and environmental improvement. In terms of economic growth, the government optimizes the investment environment through fiscal policies, reduces enterprise costs, and promotes industrial upgrading. In the energy transition, the government formulates policies and regulations to promote enterprises to develop renewable energy, reduce the use of high carbon energy, and accelerate the energy transition. In the field of environmental improvement, the government formulates strict environmental standards, strengthens supervision and law enforcement, and punishes polluting enterprises. It also increases financial investment in environmental protection for ecological restoration and pollution control.

There are many factors affecting the level of carbon emissions, like the economy, environment, energy, industrial structure, and urbanization level [10,11], and a great deal of research has been conducted on the factors affecting carbon emissions. Existing studies have concluded that government intervention is a determining factor in the process of achieving carbon emission reduction, covering policy effect evaluation, optimization studies, and cross-country studies [12]. Investigating the link between energy transformation and carbon emission efficiency can provide decision-making insights for balancing economic growth and carbon neutrality goals [13]. Controlling coal consumption, promoting clean energy, and improving energy efficiency are fundamental ways to achieve green and sustainable economic development [14]. Zhang et al. [15] argue that interaction between the industrial agglomeration level and environmental productivity affects carbon emissions. Zhang et al. [16] argue that urbanization affects production and consumption patterns, leading to energy demand change. The long-term deterioration of environmental resources urgently requires a shift from the traditional view of economic development to sustainable development [17]. Incorporating carbon emissions into the efficiency of the allocation of factors of production and exploring the low-carbon driving effect of the digital economy is a necessary and urgent research topic [18]. Weng and Xu [19] argued that China’s carbon market can restrain carbon emissions effectively. Chen et al. [20] argued that the carbon emission trading mechanism can be used as an important market-based environmental policy tool to reduce carbon emissions.

In the existing studies, although the realization of carbon neutrality and its related policies have been explored in a more comprehensive manner, there are still some shortcomings and limitations: First, when exploring the factors affecting carbon neutrality, the existing studies focus on the analysis of a variety of factors one by one and lack a systematic analysis of the interaction between different factors. This leads to a lack of in-depth understanding of the complexity and multidimensionality of carbon emission reduction, making it difficult to effectively formulate comprehensive policy measures. Second, the role of market mechanism in carbon emission reduction is mostly emphasized, but the research on the macro-control role of government financial intervention in carbon emission reduction is relatively weak. This leads to insufficient understanding of the government’s role in guiding and regulating carbon emission reduction, making it difficult to give full play to the positive role of government fiscal policy in carbon emission reduction. Finally, it fails to fully consider the dynamic changes and timeliness of the government fiscal intervention and lacks the examination of the nonlinear and lagging effects that may occur in the process of policy implementation.

Based on the above analysis, the following issues become crucial in the process of realizing carbon neutrality.

Can government fiscal intervention effectively regulate the economic system, environmental system, and energy system so as to realize an effective reduction of total carbon emissions in Jiangsu Province?

In this paper, we will explore the mechanism of the regulating effect of economic growth (EG), energy transformation (ET), and environmental improvement (EI) on carbon emission reduction under government financial intervention and analyze its regional variability in Jiangsu Province. Based on this issue, this paper combines the DPSIRM theoretical framework model and the improved entropy value method to calculate the comprehensive evaluation level of the economic, energy, and environmental systems in Jiangsu Province, based on which this paper further constructs the threshold regression model to explore the degree of influence of EG, ET, and EI on the carbon emission level under the threshold effect of different government financial interventions. Through empirical analysis, it reveals the role mechanism of government financial intervention in carbon emission reduction and provides a theoretical basis and practical reference for the formulation of more scientific and effective carbon emission reduction policies. The study determines different turning points and inflection points according to the level of government financial intervention by calculating the threshold estimation value and conducting the threshold regression analysis to portray the extent of the role of EG, ET, and EI on the carbon emission level, respectively, with the change of government financial intervention and reveal the internal mechanism of the role between systems.

Importantly, the highlights of this article and differences from previous research are as follows:

First, from a research perspective, this article takes local government fiscal intervention as an important research variable; analyzes its regulatory effect on carbon emission reduction; and studies economic growth, energy transformation, and environmental improvement as a whole to explore the impact of system coordinated development on carbon emissions. The existing research mainly focuses on the role of market mechanisms in carbon reduction [21] and mostly focuses on the analysis of the degree of impact of a single influencing factor on achieving carbon reduction, and there is insufficient research on the interaction of multiple factors and the synergistic effect of government intervention [22]. Second, in terms of research methods, the DPSIRM theoretical framework model is combined with the improved entropy method to construct a coordinated development evaluation system of economic growth, energy transformation, and environmental improvement. This reflects the innovation of multidimensional evaluation and data processing methods for indicators and fills the gap in the determination of indicator evaluation weights in existing research [23,24]. Furthermore, by constructing a threshold effect model to identify the impact of fiscal intervention on carbon reduction and exploring whether there is a significant change in the impact of fiscal intervention on carbon reduction when it reaches the critical point, most studies ignore the nonlinear effects of fiscal intervention [25] and lack research on government intervention mechanisms [26]. Furthermore, based on the research results, the moderating effect of government fiscal intervention on the impact of energy transformation, economic growth, and environmental improvement on carbon emission levels is proposed: as the degree of government fiscal intervention increases, energy transition, economic growth, and environmental governance promote carbon reduction to varying degrees, and their impact mechanisms exhibit differentiated marginal effect characteristics. Existing research mostly focuses on the overall effectiveness of fiscal intervention policies, lacking a dynamic analysis of fiscal policies and identification and exploration of the optimal intervention degree [27,28]. Figure 1 is the structural flowchart of this article.

2. Literature Review

2.1. Factors Influencing Carbon Emissions

Regarding the impact of EG, ET, and EI on carbon emissions, scholars from both domestic and international contexts have carried out extensive research. Aiming at the variable relationship between carbon emissions and economic growth, previous studies mainly centered on the environmental Kuznets curve prediction of carbon dioxide and the coordinated relationship between carbon emissions and economic growth [29,30]. Using the decoupling index to analyze the dynamic coupling relationship between economic growth and carbon emissions, the study indicates that the growth of economic scale presents a positive driving effect on the growth of carbon emissions as a whole [31]. Research on environmental improvement and carbon emissions mainly concentrated on the mechanism of environmental regulation on the effect of traded carbon emissions [32]. In response to environmental regulations, traded enterprises will increase R&D investment in energy saving and emission reduction and green development transformation, which in turn will reduce carbon emissions [33,34]. Huang et al. [35] explored whether the carbon emission trading policy would bring environmental dividends and examined its promotional effect on carbon emission reduction. There is a close link between the carbon emission trading pilot policy and energy transformation, energy utilization efficiency and renewable energy development, and it is found that emission trading pilot programs effectively reduced regional energy consumption and intensity, enhanced energy efficiency, and fostered renewable energy growth [36,37]. Extensive studies have explored the impact of EG, ET, and EI on carbon emissions, but some limitations remain. Many studies have focused on the analysis of multiple factors one by one and lacked a systematic analysis of the interaction and comprehensive impact of different factors. The assessment of the actual effect after the implementation of the policy is still insufficient, and existing research lacks empirical evidence on the policy’s impact. This article constructs an indicator system of economic growth, energy transformation, and environmental improvement, comprehensively considering the impact on carbon emission levels.

2.2. Duality of Government Intervention

With regard to the impact of government intervention on carbon emission reduction, existing studies have shown that it has two sides. On the one hand, government intervention may lead to inefficiency in market resource allocation; on the other hand, macro-issues such as market stability and financial risk still need to be solved by the government-led solution, and the realization of emission reduction targets cannot be achieved without the dual role of market mechanism and government intervention. Government financial intervention is an effective means to adjust the direction of market demand and enhance marketability [38]. In the face of increasingly severe and complex environmental problems, due to the public goods nature of resources and environment and the failure of market adjustment, the government often carries out the necessary intervention. Zhao et al. [39] found that the stronger the degree of government intervention is in the economy, the lower the degree of financial resource mismatch is between SOEs through the study of financial resource mismatch. Wang et al. [40] found that there is a nonlinear “U”-shaped relationship between government intervention and carbon emissions, and Li [41] demonstrated that the impact of government intervention on carbon emissions unfolds in a three-stage nonlinear pattern. Wu et al. analyzed the carbon reduction effect of China’s carbon market from the perspective of the synergistic effect of market mechanisms and administrative intervention and found that the carbon market has a significant carbon reduction effect, and the greater the government’s administrative intervention in the carbon market is, the stronger the carbon reduction effect of the carbon market is [42]. Yuan et al. conducted empirical tests using a system GMM model based on a centralized and decentralized institutional governance system and found that there is a “U”-shaped relationship between central vertical and local parallel environmental regulations and high-quality emission reduction development. This verifies the existence of an environmental EKC curve relationship, reflecting the nonlinear relationship between government environmental regulations and carbon emission reduction targets [43]. One of the focuses of foreign studies on government environmental improvement behavior is to focus on the slight behavior of local environmental policies in the context of environmental decentralization, which has shown that environmental decentralization will trigger a “bottom-by-bottom competition” of environmental standards among localities and face the problem of cross-regional spillovers of environmental pollution [44]. Much of the literature has focused on regional environmental improvement behavior at the local government level [45]. Despite numerous studies examining the impact of government intervention on carbon emission reduction, several limitations persist. Many studies focus only on the unilateral role of government intervention and lack in-depth analysis of the duality of government intervention and the complex impact it brings. Existing studies tend to focus on the linear relationship when analyzing the relationship between carbon emissions and government intervention and lack sufficient attention to the nonlinear relationship and threshold effect. This paper captures the nonlinear relationship and dynamic relationship of each factor on the level of carbon emissions under the threshold effect of government intervention by constructing a fixed effect model and a threshold effect model.

2.3. DPSIRM Theoretical Model

The DPSIR theoretical model, which means the Driving Force–Pressure–State–Impact–Response (DPSIR) framework, provides a structured approach for analyzing and assessing environmental issues and their progression. The model was proposed by OECD in 1993 [46] and has been widely used in environmental management, policy making, and research. It is often used in the construction of complex coupled system models, which can reflect the causal relationship between different components of the model [47]. When the DPSIR model is used to construct the indicator system, due to the interconnection between the components of driving force, pressure, state, impact, and response, it is able to carry out an overall dynamic simulation of the ecosystem in the watershed, which provides a basis for scientifically evaluating the ecological security status, indicating the applicability of the model in the study of ecological security evaluation [48,49,50,51]. The DPSIRM model emphasizes the causal linkages among social, economic, and ecological systems and is applicable to water resources security, lake ecosystem health, and human habitat security, which fits the characteristics of ecological security governance [52,53]. The causal linkage of government intervention and restoration measures as a separate management module can emphasize the coupling and synergistic effects of the natural environment, resources, and human activities and ensures the ecosystem health and system integrity of the management area [54]. The DPSIR model has shown significant advantages in environmental problem analysis and assessment, but existing research still has some limitations. Some studies focus on the construction and application of models while neglecting the analysis of the theoretical basis and limitations of the models themselves. In addition, existing research lacks evaluation of indicator weights and the effectiveness of indicator systems when using DPSIR models to construct indicator systems, which may affect the application effectiveness of the models. This article combines the DPSIRM theoretical framework model with the entropy method, taking into account the six elements of the theoretical model in indicator weights, which can more specifically reflect the relationships between each element.

3. Model Construction and Variable Selection

3.1. Construction of a Comprehensive Evaluation Index System for Coordinated Development of Economic Growth, Energy Transformation, and Environmental Improvement

This article was based on the DPSIRM theoretical model framework, analyzing and evaluating environmental issues in a structured manner and identifying key indicators for the coordinated development of EG, ET, and EI in Jiangsu Province. It used the improved entropy method to calculate indicator weights, removed outliers in the data that exceed a certain range, and then calculated the mean. It also reduced the impact of data outliers on indicator weight calculation and improves the stability and reliability of weight allocation.

First, the DPSIRM theoretical framework model was adopted to construct an indicator system from six perspectives. The specific functions are shown in Figure 2. Among them, the “driving force” factor is the source that affects economic development, usually directly or indirectly causing changes in pressure factors. It is the most fundamental driving force for promoting economic development, mainly manifested in the aspect of driving economic growth. The “pressure” factor refers to the changes caused by the direct driving force of energy transformation and environmental improvement, usually with negative effects. The “state” factor is the actual state exhibited under the combined force of driving forces and pressures. This study mainly included the economic growth state, energy transformation state, and environmental improvement state. The “impact” factor is the combined effect of driving force, pressure, and state, presenting a state that promotes or inhibits the system. This study considered the impact on economic growth, energy transformation, and environmental improvement. The “response” factor is a response measure taken to improve or adapt to the current state and achieve sustainable development, which is a feedback on driving forces and pressures. The “management” factor refers to the proactive intervention and measures taken by humans to promote the coordinated development of the system, mainly necessary measures taken by the government.

Based on the connotations and characteristics of economic growth (EG), energy transformation (ET), and environmental improvement (EI), the existing literature’s indicator system was summarized and organized. Combined with the DPSIRM model, an indicator system was constructed from six perspectives. At the same time, considering the coordinated development goals and actual development status of Jiangsu Province, we selected the following indicators for measurement. Among them, economic growth included 10 secondary indicators, energy transformation included 8 secondary indicators, and environmental improvement included 10 secondary indicators, reflecting the level of EG, ET, and EI in Jiangsu Province.

Adopting the improved entropy method to measure the comprehensive evaluation level of EG, ET, and EI in Jiangsu mainly included the following steps.

Step 1: Eliminate the influence of different dimensions by standardizing each indicator.

Positive indicators:

$$U_{ij}={\displaystyle \frac{X_{ij}-\min X_{ij}}{\max X_{ij}-\min X_{ij}}}$$

(1)

Negative indicator:

$$U_{ij}={\displaystyle \frac{\max X_{ij}-X_{ij}}{\max X_{ij}-\min X_{ij}}}$$

(2)

Step 2: Introduce the DPSIRM-improved entropy weight coefficient $\delta kj$, and indicate the importance of the theoretical factor $k$ corresponding to the $j$ indicator, satisfying the following:

$${\displaystyle \sum _{k=1}^L{\delta }_{kj}}=1$$

(3)

Step 3: Calculate the proportion of indicator $j$ for the year $i$:

$$y_{ij}=U_{ij}/{\displaystyle \sum _{i=1}^m{U}_{ij}}$$

(4)

Step 4: Calculate the information entropy of the indicator $j$:

$$K={\displaystyle \frac{1}{\ln m}}$$

(5)

$$e_j=-K{\displaystyle \sum _{i=1}^m{y}_{ij}\ln y_{ij}}$$

(6)

Step 5: Calculate the weighted entropy value of each indicator $e_j^{\prime }$:

$$e_j^{\prime }=wj\times ej$$

(7)

Step 6: Calculate the weighted difference coefficient for each indicator $g_j^{\prime }$:

$$g_j^{\prime }=1-e_j^{\prime }$$

(8)

Step 7: Calculate the weight of the indicators $w_j^{\prime }$:

$$w_j^{\prime }={\displaystyle \frac{g_j^{\prime }}{{\displaystyle {\sum }_{j=1}^n{g}_j^{\prime }}}}$$

(9)

Step 8: Calculate the comprehensive score of each evaluation object $S_i$:

$$S_i={\displaystyle \sum _{j=1}^n{w}_j^{\prime }}\cdot U_{ij}={\displaystyle \sum _{j=1}^n\frac{\left(1-{\displaystyle \sum _{k=1}^L{\delta }_{kj}}\cdot e_j\right)}{{\displaystyle {\sum }_{j=1}^n\left(1-{\displaystyle \sum _{k=1}^L{\delta }_{kj}}\cdot e_j\right)}}}\cdot U_{ij}$$

(10)

In the above equation, $X_{ij}$ represents the value of the indicator $j$ in the year $i$, $\max X_{ij}$ and $\min X_{ij}$ represent the maximum and minimum values of each evaluation indicator, $y_{ij}$ represents the weight of the indicator $j$ in the year $i$, $U_{ij}$ represents the standardized indicator, $n$ represents the total number of evaluation indicators in each year, $m$ indicate the number of years, $e_j$ represents the information entropy of the indicator $j$, and $w_j^{\prime }$ represents the weight of the indicator $j$. Check if the weights are logical and reflect the importance of the indicators. Finally, apply the calculated weights to the indicator data and calculate the comprehensive evaluation level of the indicators $S_i$. In order to simplify subsequent calculations in improving the entropy method, the indicator weight ${\delta }_{kj}$ is uniformly set to 1/5.

Table 1. Construction and weight of coordinated development indicators for economic growth, energy transformation, and environmental improvement in Jiangsu.

Primary Indicator	Theory	Secondary Indicators	Indicator Attribute	Unit	Weight
$EG$	driving force	AGDP	+	CNY	0.0841
		Urban population ratio	+	%	0.0979
	state	Per capita income	+	CNY	0.0817
		Tertiary industry growth rate	+	%	0.0858
	impact	Per capita ratio	+	%	0.0766
		Per capita disposable income	+	CNY	0.1204
		Per capita living consumption expenditure	+	CNY	0.0862
		Social labor productivity	+	%	0.188
	response	Patent authorization quantity	+	unit	0.091
	management	R&D budget expenditure	+	100 m RMB	0.0883
$ET$	pressure	Energy consumption per-unit GDP	-	10 kt /100 m Y	0.1001
		Per capita domestic energy consumption	-	10 kt/10 k people	0.1788
	state	Industrial electricity consumption	-	GWh	0.1521
		Social electricity consumption	-	GWh	0.1668
		Total energy consumption growth	-	%	0.0997
	impact	Elasticity of energy consumption	+	%	0.2045
	response	Energy utilization rate	+	%	0.047
	management	Investment in energy technology R&D	+	100 m RMB	0.051
$EI$	pressure	Industrial SO2 emissions	-	ton	0.0838
		Industrial wastewater discharge	-	stere	0.1161
		Industrial smoke and dust emissions	-	ton	0.1464
	state	Urban precipitation	+	mm	0.1037
		Total water resources	+	Gm3	0.1025
	impact	Green coverage rate	+		0.0616
	response	Utilization rate of industrial solid waste	+	%	0.1060
		Urban sewage treatment rate	+	%	0.0810
		Harmless treatment of domestic garbage	+	%	0.0768
	management	Environmental protection budget expenditure	+	100 m RMB	0.1223

shows the construction of the evaluation index system and the obtained weights.

3.2. Calculation of Carbon Emissions

This article used the terminal energy consumption method to calculate the carbon emission level of Jiangsu from 2010 to 2020. It was based on the energy balance sheet terminal energy consumption to estimate the carbon dioxide emissions of the terminal. The specific formula is as follows:

$${CE}_{energy}={\displaystyle \sum _i{\displaystyle \sum _j{E}_{ij}}}·{\alpha }_j·{\beta }_j·(44/12)$$

(11)

In the formula, $CE$ represents the total carbon dioxide emissions of a certain region. The subscript $i$ represents the industry (i = 1, 2, …, 6, respectively, representing agriculture, industry, construction, transportation and warehousing, postal services, and other industries). The subscript $j$ represents the types of energy (the classification of energy types varies in the annual energy statistical yearbooks); $E_{ij}$ represents the physical quantity of $j$ types of energy consumed by industry $i$; ${\alpha }_j$ is the conversion coefficient to standard coal for $j$ types of energy; $\beta j$ is the carbon emission coefficient for $j$ types of energy; and 44/12 is the conversion coefficient for carbon to carbon dioxide.

After calculation, the trend of changes in carbon emission levels in Jiangsu from 2010 to 2020 is shown in Figure 3.

From Figure 3, it can be seen that the carbon emission level in Jiangsu Province has generally shown an upward trend in the past decade, increasing from an initial 20.2 to around 22.8. Among it, the growth rate was relatively fast from 2010 to 2013 but significantly slowed down from 2013 to 2020, and there was a brief downward trend from 2015 to 2016 and from 2017 to 2018. Overall, although the carbon emissions in Jiangsu Province are on the rise, the growth rate is slowing down. This demonstrates the efforts of Jiangsu Province in achieving sustainable development and addressing climate change. Through policy guidance, industrial restructuring, and technological innovation, Jiangsu Province is moving towards a lower carbon and more environmentally friendly direction.

3.3. Variable Selection and Data Sources

(1)

Dependent variable and explanatory variable.

The dependent variable is the level of carbon emissions ($CE$), represented by the total carbon emissions. The carbon emission data were sourced from the Energy Statistical Yearbook and the Energy Statistical Yearbook of various prefecture level cities. The specific calculation method is described in Section 3.2 above. Economic growth ($EG$), energy transformation ($ET$), and environmental improvement ($EI$) are the explanatory variables, represented by the comprehensive development level of each indicator $S_i$ calculated in Section 3.1 above.

(2)

Control variables and threshold variables.

Referring to existing research and experience, control variables such as openness, industrial structure, urbanization level, technological progress, informatization level, and government financial intervention were selected, with the threshold variable being government financial intervention. Specifically, the degree of openness ($OD$) is expressed in terms of export value, measured in billions of CNY. The industrial structure ($IS$) is represented by the ratio of $GDP$ added value of the tertiary industry. The level of urbanization ($UL$) is measured by the urban population’s share of the total population. Technological progress ($TP$) is the ratio of capital stock to labor force. The level of informatization ($IL$) is the total expenditure on transportation and communication. The government fiscal intervention ($GOV$) adopts the ratio of general public budget expenditure to gross domestic product.

(3)

Data source.

Considering the accuracy and availability of data, this article took panel data from various cities in Jiangsu from 2010 to 2020 as samples. The data mainly came from relevant statistical materials such as Jiangsu Statistical Yearbook, Urban Statistical Yearbook, Energy Statistical Yearbook, etc. Some missing data points were filled in using interpolation methods, while the emission factors were sourced from officially published data, including the “Guidelines for Provincial Greenhouse Gas Emission Inventories (Trial)” and others.

Table 2. Descriptive statistics of variables.

Variable	Observations	Mean Value	SD.	Min Value	Max Value
$CE$	143	1.701	0.354	0.786	2.374
$EG$	143	4.118	2.627	0.01	10.006
$ET$	143	4.686	1.564	1.641	8.412
$EI$	143	4.733	2.031	1.402	9.507
$OD$	143	53.748	59.705	2	333
$IS$	143	5.654	4.178	−24.454	19.179
$UL$	143	65.734	9.482	48.3	86.8
$TP$	143	6.559	1.701	3.581	10.143
$IL$	143	2.832	1.319	1.095	6.324
$GOV$	143	1.241	0.321	0.8	2.067

reports descriptive data for variables $CE$, $EG$, $ET$, and $EI$ with a total observation value of 143 for each variable. The average carbon emission level was 1.701, with a standard deviation of 0.354, and the minimum and maximum data ranged from 0.786 to 2.374. The levels of EG, ET, and EI were all positive averages, with the smallest standard deviation for energy transformation and the largest standard deviation for economic growth. Overall, the average level of carbon emissions was 1.701, and EG, ET, and EI were all positive averages, indicating that efforts in economic development and environmental protection were yielding results. Although the volatility of economic growth indicated differences between different regions, the relative consistency of energy transformation and positive progress in environmental improvement provided a good foundation for future sustainable development. In order to further reduce carbon emissions, various regions should continue to facilitate the enhancement of energy structure optimization, enhance the green transformation of the economy, implement stricter environmental improvement measures, and achieve more efficient resource utilization and sustainable development goals.

3.4. Construction of Econometric Models

To assess the impact of EG, ET, and EI on carbon emission reduction in Jiangsu Province, respectively, under the regulation effect of government fiscal intervention, this paper adopted the method of constructing a threshold effect model to realize it. First, any threshold variable was randomly selected as the threshold, the data were divided into two intervals, and then the parameter values of these two intervals were estimated by OLS method. The total residuals of the sum of squares of the two intervals were calculated, the sum of squares of the residuals was recorded, different thresholds were selected, and the sum of squares of the residuals corresponding to the thresholds was recorded. Successive repetitive comparisons of the residual sum of squares were made, and the threshold value corresponding to its minimum value was the best threshold estimate. Finally, the existence and veracity of the thresholds were tested using hypothesis testing, by which the threshold effect of the model was determined.

(1)

Fixed panel model.

First of all, in order to initially verify the degree of influence of EG, ET, and EI on the level of carbon emissions in Jiangsu Province, respectively, it is essential to establish a panel effect model, and through the results of the Hausman test, this paper finally selected the panel fixed effect model regression. Meanwhile, multicollinearity tests were conducted on the explanatory variables to ensure the accuracy of parameter estimation and the effectiveness of the model. The formulas are as follows.

$${CE}_{it}={\alpha }_0+{\alpha }_1{EG}_{it}+{\alpha }_{2}E{G}_{it}^2+{\alpha }_3{Z}_{it}+{\mu }_i+{\delta }_t+{\epsilon }_{it}$$

(12)

$${CE}_{it}={\beta }_0+{\beta }_1{ET}_{it}+{\beta }_{2}E{T}_{it}^2+{\beta }_3{Z}_{it}+{\mu }_i+{\delta }_t+{\epsilon }_{it}$$

(13)

$${CE}_{it}={\chi }_0+{\chi }_{1}E{I}_{it}+{\chi }_{2}E{I}_{it}^2+{\chi }_3{Z}_{it}+{\mu }_i+{\delta }_t+{\epsilon }_{it}$$

(14)

Among them, ${CE}_{it}$ is the carbon emissions of each city $i$ in different years $t$; ${\alpha }_0$ represents the intercept term; ${EG}_{it}$, ${ET}_{it}$, and ${ET}_{it}$ are the economic growth level, energy transformation level, and environmental improvement level of each city $i$ in different years $t$; and $Z_{it}$ is a series of control variables, including opening degree, industrial structure, urbanization level, technological progress, government financial intervention, and informatization level. ${\mu }_i$ is an individual fixed effect, ${\delta }_t$ is a time fixed effect, and ${\epsilon }_{it}$ is a random perturbation term.

(2)

Construction of threshold effect model.

On the basis of the fixed effects model, this article used Hansen’s panel threshold regression model for testing [55], with Jiangsu Provincial Government fiscal intervention as the threshold variable, to examine the varying degrees of impact and trends of Jiangsu EG, ET, and EI on carbon emission levels under different government fiscal interventions. The government, as the market economy’s main body, may differently impact carbon emissions through policy enforcement across economic, energy, and environmental systems. This article predicted that carbon emission reduction with loose or strict levels of government fiscal intervention may be detrimental to the optimization and improvement of the economic, energy, and environmental systems, thereby affecting the reduction of carbon emissions. Appropriate government fiscal intervention can be more effective, as there may be a nonlinear relationship between the economic, energy, and environmental systems and carbon emission mitigation under government fiscal intervention.

The advantage of this model is that when studying the nonlinear relationship between independent and dependent variables, users do not need to provide nonlinear equations. On the contrary, the number and threshold values are determined by the sample data. This method avoids errors caused by manually dividing samples, divides intervals based on thresholds, and compares the differences in regression coefficients after internal grouping. Threshold regression is a nonlinear econometric model that estimates threshold values based on the inherent causal laws of sample data and tests whether there are significant differences between sample groups divided by threshold values. According to the basic principle of this model, the threshold variable is allowed to be the explanatory variable or other control variables in the model. Therefore, for the fixed effects model mentioned above, a threshold effect model is designed as follows:

$$\begin{array}{ll}C{E}_{it}=& {\delta }_0+{\delta }_{1}E{G}_{it}\cdot I\left(GO{V}_{it}\le \gamma \right)+{\delta }_{2}E{G}_{it}\cdot I(GO{V}_{it}>\gamma )+\hfill \\ & {\delta }_{3}O{D}_{it}+{\delta }_{4}I{S}_{it}+{\delta }_{5}U{L}_{it}+{\delta }_{6}T{P}_{it}+{\delta }_{7}I{L}_{it}+{\delta }_{8}GO{V}_{it}+{\epsilon }_{it}\hfill \end{array}$$

(15)

$$\begin{array}{ll}C{E}_{it}=& {\phi }_0+{\phi }_{1}E{T}_{it}\cdot I\left(GO{V}_{it}\le \theta \right)+{\phi }_{2}E{T}_{it}\cdot I(GO{V}_{it}>\theta )+\hfill \\ & {\phi }_{3}O{D}_{it}+{\phi }_{4}I{S}_{it}+{\phi }_{5}U{L}_{it}+{\phi }_{6}T{P}_{it}+{\phi }_{7}I{L}_{it}+{\phi }_{8}GO{V}_{it}+{\epsilon }_{it}\hfill \end{array}$$

(16)

$$\begin{array}{ll}C{E}_{it}=& {\sigma }_0+{\sigma }_{1}E{I}_{it}\cdot I\left(GO{V}_{it}\le \omega \right)+{\sigma }_{2}E{I}_{it}\cdot I(GO{V}_{it}>\omega )+\hfill \\ & {\sigma }_{3}O{D}_{it}+{\sigma }_{4}I{S}_{it}+{\sigma }_{5}U{L}_{it}+{\sigma }_{6}T{P}_{it}+{\sigma }_{7}T{L}_{it}+{\sigma }_{8}GO{V}_{it}+{\epsilon }_{it}\hfill \end{array}$$

(17)

The meaning of the corresponding variables in the formula remains unchanged. $I(•)$ represents a demonstrative function. Using $EG$, $ET$, and $EI$ as the core explanatory variables; $GOV$ as the threshold variable; and $\gamma$, $\theta$, and $\omega$ as specific threshold values, the sample can be divided into two groups based on the threshold values.

On the basis of the single-threshold model, we can continue to examine the situation where there are multiple threshold values, such as the construction of the dual-threshold model as follows:

$$\begin{array}{ll}C{E}_{it}=& {\delta }_0+{\delta }_{1}E{G}_{it}\cdot I\left(GO{V}_{it}\le {\gamma }_1\right)+{\delta }_{2}E{G}_{it}\cdot I({\gamma }_1<GO{V}_{it}\le {\gamma }_2)+{\delta }_{3}E{G}_{it}\cdot I(GO{V}_{it}>{\gamma }_2)+\hfill \\ & {\delta }_{4}O{D}_{it}+{\delta }_{5}I{S}_{it}+{\delta }_{6}U{L}_{it}+{\delta }_{7}T{P}_{it}+{\delta }_{8}I{L}_{it}+{\delta }_{9}GO{V}_{it}+{\epsilon }_{it}\hfill \end{array}$$

(18)

$$\begin{array}{ll}C{E}_{it}=& {\phi }_0+{\phi }_{1}E{T}_{it}\cdot I\left(GO{V}_{it}\le {\theta }_1\right)+{\phi }_{2}E{T}_{it}\cdot I({\theta }_1<GO{V}_{it}\le {\theta }_2)+{\phi }_{3}E{T}_{it}\cdot I(GO{V}_{it}>{\theta }_2)+\hfill \\ & {\phi }_{4}O{D}_{it}+{\phi }_{5}I{S}_{it}+{\phi }_{6}U{L}_{it}+{\phi }_{7}T{P}_{it}+{\phi }_{8}I{L}_{it}+{\phi }_{9}GO{V}_{it}+{\epsilon }_{it}\hfill \end{array}$$

(19)

$$\begin{array}{ll}C{E}_{it}=& {\sigma }_0+{\sigma }_{1}E{I}_{it}\cdot I\left(GO{V}_{it}\le {\omega }_1\right)+{\sigma }_{2}E{I}_{it}\cdot I({\theta }_1<GO{V}_{it}\le {\omega }_2)+{\sigma }_{3}E{I}_{it}\cdot I(GO{V}_{it}>{\omega }_2)+\hfill \\ & {\sigma }_{4}O{D}_{it}+{\sigma }_{5}I{S}_{it}+{\sigma }_{6}U{L}_{it}+{\sigma }_{7}T{P}_{it}+{\sigma }_{8}I{L}_{it}+{\sigma }_{9}GO{V}_{it}+{\epsilon }_{it}\hfill \end{array}$$

(20)

Among them, ${\gamma }_1<{\gamma }_2$, ${\theta }_1<{\theta }_2$, and ${\omega }_1<{\omega }_2$, the calculation process of the dual-threshold model is similar to that of a single threshold; that is, in the case where the first threshold value is fixed, the second threshold value is searched for. The principle of the three threshold model is similar and will not be further elaborated.

4. Empirical Results and Analysis

4.1. Fixed Effects Regression Model Analysis

After conducting the Hausman test on the benchmark OLS model, the p-values were all 0.000, stating that the null hypothesis of random effects was rejected with a significance level of 1%, leading to the acceptance of the alternative hypothesis of fixed effects. In this study, the logarithm of the variables was taken to eliminate the influence of heteroscedasticity. The results are shown in

Table 3. Fixed effects regression results.

Explanatory Variable	(1)	(2)	(3)
	$\mathit{E}\mathit{G}$	$\mathit{E}\mathit{T}$	$\mathit{E}\mathit{I}$
$EG$	−0.173 **
	(0.085)
$E{G}^2$	0.0028 *
	(0.0023)
$ET$		−0.243 **
		(0.120)
$E{T}^2$		−0.0030 ***
		(0.0012)
$EI$			−0.153 *
			(0.111)
$E{I}^2$			−0.0003
			(0.0013)
$OD$	0.018 ***	0.022***	0.019 ***
	(0.007)	(0.007)	(0.007)
$IS$	0.067 *	0.053	0.053
	(0.050)	(0.049)	(0.049)
$UL$	0.262 ***	0.042 ***	0.246 ***
	(0.059)	(0.058)	(0.058)
$TP$	−0.543 **	−0.604 **	−0.462 *
	(0.272)	(0.273)	(0.282)
$IL$	−0.222	−0.463 *	−0.300
	(0.366)	(0.356)	(0.366)
$GOV$	−0.2099 ***	−0.2975 ***	−0.2013 **
	(0.0802)	(0.0806)	(0.0864)
Constant	0.5917 **	0.9765 ***	0.6631 **
	(0.2825)	(0.2991)	(0.2807)
$R^2$	0.6591	0.6588	0.6535

Notes: ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively; the standard error is in parentheses.

.

As shown in Table 3, economic growth has a significant inhibitory effect on carbon emissions at the 5% level, energy transformation results in a marked decrease in carbon emissions at the 5% level, and environmental improvement has a significant inhibitory effect on carbon emissions at the 10% level. It follows that EG, ET, and EI have all contributed to carbon reduction to a certain extent. The squared term of economic growth is significant at the 10% level and carries a positive sign, suggesting that the relationship between economic growth and carbon emissions follows an upward-opening parabolic trajectory or, in other words, exhibits a “U”-shaped pattern. This indicates that after reaching a certain critical value of economic growth, the increase in economic growth will promote carbon emissions, that is, suppress carbon reduction. In the initial stage of economic growth, as the economic scale gradually expands, factors such as industrial restructuring and improved resource allocation efficiency lead to a reduction in carbon emissions. However, when a certain critical value is reached, economic growth actually promotes carbon emissions. This is because in the stage of rapid economic growth, in order to pursue economic benefits, environmental protection is often neglected. Large-scale industrial construction and production activities can lead to a sharp increase in energy consumption, which suppresses the effect of carbon reduction. Similarly, the square term of energy transformation is significant at the 1% level and has a negative sign. The relationship between energy transformation and carbon emissions shows an inverted “U” shape, revealing that after reaching a tipping point, the improvement of energy transformation will promote carbon reduction. In the early stage of energy transformation, due to investment in technology research and infrastructure construction, the speed of transformation may be relatively slow, and the impact of energy structure adjustment on carbon emissions is not yet significant. With the advancement of energy transition, further improvement in the level of energy transition will greatly promote carbon reduction. The square term of environmental improvement is not significant. This indicates that the impact of environmental governance on carbon emissions does not show a significant nonlinear relationship. Overall, this result indicates that under the current level of environmental governance and relevant conditions, continuously strengthening environmental governance is a reliable and effective strategy for reducing carbon emissions.

In terms of control variables, openness, industrial structure, and the urbanization level show a positive relationship with carbon emissions and are significant at the 1%, 10%, and 1% levels, respectively. This means that higher openness and urbanization levels may lead to higher energy consumption and emissions, especially in rapidly developing cities where industrial structure adjustments fail to keep up in a timely manner, resulting in high resource consumption and increased emissions. There is a negative relationship between the information technology level, technological progress, and government financial intervention, which are significant at the 5%, 10%, and 1% levels, respectively. This indicates that the improvement of technological progress and the information technology level can reduce carbon emissions, and government financial support can also help promote the application of low-carbon technologies and the implementation of environmental improvement.

In order to ensure the reliability of the benchmark regression results, this article uses three methods for robustness testing, as shown in

Table 4. Robustness analysis (with added control variables).

Explanatory Variable	(1)	(2)	(3)
	$\mathit{E}\mathit{G}$	$\mathit{E}\mathit{T}$	$\mathit{E}\mathit{I}$
$EG$	−0.018 **
	(0.009)
$ET$		−0.025 **
		(0.012)
$EI$			−0.019
			(0.012)
$OD$	0.002 **	0.002 ***	0.002 **
	(0.001)	(0.001)	(0.001)
$IS$	0.006	0.005	0.004
	(0.005)	(0.005)	(0.005)
$UL$	0.028 ***	0.026 ***	0.028 ***
	(0.007)	(0.007)	(0.007)
$TP$	−0.051 *	−0.058 **	−0.04
	(0.028)	(0.028)	(0.029)
$IL$	−0.035	−0.06	−0.05
	(0.044)	(0.043)	(0.043)
$HC$	−0.004	−0.004	−0.006
	(0.007)	(0.007)	(0.007)
$GOV$	−0.217 ***	−0.307 ***	−0.204 **
	(0.081)	(0.083)	(0.086)
Constant	0.506	0.899 ***	0.513
	(0.324)	(0.33)	(0.332)
$R^2$	0.660	0.660	0.655

Notes: ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively; the standard error is in parentheses.

,

Table 5. Robustness analysis (replacing the dependent variable).

Explanatory Variable	(1)	(2)	(3)
	$\mathit{E}\mathit{G}$	$\mathit{E}\mathit{T}$	$\mathit{E}\mathit{I}$
$EG$	−0.022 ***
	(0.003)
$ET$		−0.015 ***
		(0.005)
$EI$			−0.031 ***
			(0.004)
$OD$	−0.001 ***	−0.001 ***	−0.001 ***
	(0.001)	(0.001)	(0.001)
$IS$	0.003 *	0.006 ***	0.004 **
	(0.002)	(0.002)	(0.002)
$UL$	0.002	0.003	0.004 *
	(0.002)	(0.002)	(0.002)
$TP$	−0.002	−0.004	−0.019 *
	(0.01)	(0.011)	(0.01)
$IL$	0.036 ***	0.056 ***	0.037 ***
	(0.014)	(0.015)	(0.013)
$GOV$	0.04	0.065 **	−0.011
	(0.03)	(0.034)	(0.03)
Constant	7.509 ***	7.483 ***	7.476 ***
	(0.105)	(0.125)	(0.099)
$R^2$	0.539	0.418	0.579

Notes: ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively; the standard error is in parentheses.

and

Table A1. Robustness analysis (replacing the independent variable).

	(1)	(2)	(3)
Variables	$\mathit{E}\mathit{G}$	$\mathit{E}\mathit{T}$	$\mathit{E}\mathit{I}$
$EG$	−0.0460 **
	(0.0196)
$ET$		−0.243 **
		(0.120)
$EI$			−0.153
			(0.111)
$OD$	0.0017 **	0.0022 ***	0.0020 ***
	(0.0007)	(0.0007)	(0.0007)
$IS$	0.0075	0.0053	0.0053
	(0.0050)	(0.0049)	(0.0049)
$UL$	0.0300 ***	0.0242 ***	0.0246 ***
	(0.0062)	(0.0058)	(0.0059)
$TP$	−0.0389	−0.0604 **	−0.0462
	(0.0279)	(0.0273)	(0.0282)
$IL$	0.0280	−0.0463	−0.0300
	(0.0458)	(0.0356)	(0.0366)
$GOV$	−2.977 ***	−2.975 ***	−2.013 **
	(0.794)	(0.806)	(0.864)
Constant	0.532 *	0.977 ***	0.663 **
	(0.286)	(0.299)	(0.281)
Observations	143	143	143
R-squared	0.662	0.659	0.653

Notes: ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively; the standard error is in parentheses.

(Appendix A). The first approach is to increase the control variable by adding the variable of human capital ($HC$) on the basis of the original variable. The second method is to replace the dependent variable. The original carbon emission calculation used the urban carbon emissions used in the CEADs national carbon accounting database, and now, the replacement method is used to calculate the carbon emissions. Following empirical validation, the coefficient characteristics of the outcomes align closely with the initial regression findings, validating the benchmark regression model.

4.2. Threshold Effect Regression Analysis

(1)

Unit root test.

To ensure the reliability of threshold effect regression, we first conduct panel unit root tests to determine if the dataset is a stationary sequence prior to empirical analysis. The LLC test is shown as follows:

$$y_{it}=u_i+{{\beta }^{\prime }}_1{x}_{it}I(q_{it}\le \gamma )+{{\beta }^{\prime }}_2{x}_{it}I(q_{it}>\gamma )+e_{it}$$

(21)

where $I(•)$ is the indicative function, $q_{it}$ is the threshold variable, and $\gamma$ is the threshold value.

$$y_{it}=\left\{\begin{array}{ll}{u}_i+{{\beta }^{\prime }}_1{x}_{it}+e_{it},& q_{it}\le \gamma \hfill \\ u_i+{{\beta }^{\prime }}_2{x}_{it}+e_{it}& q_{it}>\gamma \hfill \end{array}\right.$$

(22)

The regression equation expression differs for each interval. Sample values are classified based on the threshold-defined intervals. The coefficients are compared after regression to analyze the changes in the relationships. The principle of dual threshold and triple threshold is the same as above.

In the LLC unit root test, the p-value is zero, meeting the significance level, and the data are stable. The unit root test passes. Although the time span of the data is not large, unit root tests are still conducted to avoid the phenomenon of “spurious regression”. The homogeneous panel unit root test (LLC) method is selected to observe whether each variable has same order cointegration. As shown in

Table 6. Unit root inspection(LLC).

Variable	LLC Test
$CE$	−4.4852 ***
$EG$	−5.9924 ***
$ET$	−5.4258 ***
$EI$	−4.9214 ***
$OD$	−5.6539 ***
$IS$	−6.4971 ***
$UL$	−2.5091 ***
$TP$	−5.3368 ***
$IL$	2.9885 *
$GOV$	−2.8661 ***

Notes: *** and * indicate significance at the 1% and 10% levels, respectively.

, all of the aforementioned methods fail to accept the null hypothesis of no unit root, indicating that all variables are horizontally stationary.

(2)

Estimation of threshold value.

For any threshold value $\gamma$, the estimated values of each parameter can be obtained by summing the squared residuals:

$$S_1(\gamma )=e_i{(\gamma )}^{\prime }{e}_i(\gamma )$$

(23)

The optimal threshold value $\widehat{\gamma }$ should be set to minimize $S_1(\gamma )$ among all residual sum of squares, that is, $\widehat{\gamma }=\mathrm{argmin}{S}_1(\gamma )$. After determining the threshold number, this article further searches and tests the threshold values.

Table 7. Threshold estimation results.

Explanatory Variable	Threshold Variable	Threshold Value		95% Confidence Interval
$EG$	$GOV$	${\gamma }_1$	1.1145	(1.1120, 1.1254)
		${\gamma }_2$	1.5295	(1.3522, 1.5309)
$ET$	$GOV$	${\theta }_1$	1.028	(0.9876, 1.0283)
		${\theta }_2$	1.9085	(1.8254, 1.9124)
$EI$	$GOV$	${\omega }_1$	1.1145	(1.1075, 1.1167)
		${\omega }_2$	1.5309	(1.5145, 1.5514)

presents the threshold estimation results, mainly reporting the parameters and corresponding 95% confidence intervals of the single and double-threshold model with government financial intervention, proving the validity of the threshold values.

(3)

Significance test.

We then determine whether the threshold value for inspection truly exists. The purpose of significance testing for threshold regression models is to examine whether the estimated parameters of the model are significantly different among sample groups divided by threshold values. A null hypothesis is constructed with no threshold value as $H_0:{\beta }_1={\beta }_2$, and at the same time, LM statistics are constructed to perform statistical tests on the null hypothesis.

$$F=n{\displaystyle \frac{S_0-S_n(\widehat{\gamma })}{S_n(\widehat{\gamma })}}$$

(24)

$S_0$ is the sum of squared residuals under the null hypothesis, and $S_n$ is the sum of squared residuals under the threshold result. Under the null hypothesis $H_0$, the coefficient ${\beta }_1={\beta }_2$ causes the system of equations to degenerate into a single linear regression equation, meaning there is no threshold effect. On the contrary, it means that ${\beta }_1$ and ${\beta }_2$ will have different effects in the two intervals. The inspection results are shown in

Table 8. Threshold effect test results.

Explaining Variable	Threshold Variable	Number	BS Frequency		F-Value	p-Value	10% Critical Level	5% Critical Level	1% Critical Level
$EG$	$GOV$	Single threshold	500	18.98	0.000		16.638	16.638	16.638
		Double threshold	500	11.39	0.000		11.032	11.032	11.032
$ET$	$GOV$	Single threshold	500	12.77	0.000		5.788	5.788	5.788
		Double threshold	500	10.36	0.000		7.744	7.744	7.744
$EI$	$GOV$	Single threshold	500	17.17	0.000		15.408	15.408	15.408
		Double threshold	500	15.25	0.000		14.699	14.699	14.699

.

The Hansen test utilizes the bootstrap method to obtain a first-order asymptotic distribution and generate a p-value. The result indicates a significant threshold effect.

The above results indicate that there is a significant threshold effect on carbon emission reduction EG, ET, and EI under government financial intervention.

(1)

When economic growth is the explanatory variable, the single- and double-threshold F-value all passed the 1% significance test, and the triple threshold did not pass the test, indicating that with changes in government financial intervention intensity, economic growth has a double-threshold effect on carbon reduction in Jiangsu Province.

(2)

When energy transformation is the explanatory variable, the single-threshold F-test was statistically significant at the 1% level, and the double threshold passed the test, indicating that energy transformation has a dual-threshold effect on carbon reduction in Jiangsu Province with changes in government financial intervention intensity.

(3)

When environmental improvement is the explanatory variable, the single- and double-threshold F-value all passed the 1% significance test, and the three thresholds did not pass the test, indicating that environmental improvement has a double-threshold effect on carbon reduction in Jiangsu with changes in government financial intervention intensity.

Based on the test results of Table 7 and Table 8, the estimated value and 95% confidence interval of government fiscal intervention reflect the threshold results. The results confirmed that there is a double-threshold effect of Jiangsu provincial government fiscal intervention on economic growth and carbon emission mitigation, with F-tests of 18.98 and 11.39, respectively. At 1%, the estimated values of 1.1145 and 1.5295 are significant. There is a double-threshold effect of government fiscal intervention on energy transformation and carbon reduction, with F-tests of 12.77 and 10.36, respectively. At 1%, the estimated values of 1.028 and 1.9085 are significant. There is a double-threshold effect of government fiscal intervention on environmental improvement and carbon reduction, with F-tests of 17.17 and 15.25, respectively. At 1%, the estimated values of 1.1145 and 1.5309 are significant. The financial intervention of the Jiangsu provincial government exhibits a dual-threshold effect on the relationship between EG, ET, EI, and carbon reduction. Through these test results, this demonstrates that the impact of fiscal policy on carbon emission mitigation is nonlinear at different intensities, emphasizing the importance of moderate fiscal intervention. This discovery provides important basis for policy-making and suggests that different threshold effects should be considered when implementing fiscal policies to optimize carbon reduction strategies.

This discovery is not only reflected in research in Jiangsu Province, but also internationally. The United States adjusts the carbon emission social cost index to influence the intervention of fiscal policy on carbon reduction. After the index is raised, the policy becomes stricter, and different industries have different reactions. When fiscal intervention is strengthened, there may be a sudden change or acceleration in the carbon reduction effect, reflecting a threshold effect. This is consistent with the nonlinear impact of changes in fiscal intervention intensity on carbon reduction in Jiangsu Province’s research, emphasizing the need to pay attention to moderation [56]. The EU uses fiscal measures such as ETS and the Social Climate Fund to promote carbon reduction. The impact of changes in fiscal intervention intensity on carbon reduction varies among different industries. For example, the carbon reduction effect of energy intensive industrial sectors may be more significant when ETS policies are strengthened, showing nonlinear characteristics, similar to the threshold effect of fiscal intervention in Jiangsu Province’s research, highlighting the importance of moderate intervention [57]. The Japanese government promotes energy conservation in government agencies through legislation and financial support. The change in financial support intensity results in different stages of energy-saving effects. The initial effect gradually increases with increased investment, and after reaching a certain level, it may significantly improve. At certain nodes, the carbon reduction effect changes significantly, reflecting the threshold effect of fiscal intervention on carbon reduction. This indicates that the government should reasonably grasp the intervention level based on actual conditions to achieve the best carbon reduction effect [58].

(4)

Confidence interval test.

When it is determined that a variable has a threshold effect, the confidence interval should be further determined. The bootstrap method is used to judge by the likelihood ratio LR statistic. To test the null hypothesis $H_0:\gamma ={\gamma }_0$, the likelihood ratio statistic is:

$$LRn({\gamma }_0)=n{\displaystyle \frac{S_0(\gamma )-S_n(\widehat{\gamma })}{S_n(\widehat{\gamma })}}$$

(25)

(a)

Dual thresholds for economic growth;

(b)

Dual thresholds for energy transformation;

(c)

Dual thresholds for environmental improvement.

The horizontal axis of Figure 4, Figure 5 and Figure 6 represents the threshold value of government financial intervention, and the vertical axis represents the likelihood ratio function value. According to Hansen’s likelihood ratio test formula, the null hypothesis of a consistent threshold estimate is rejected when the LR statistic surpasses its critical value.

When economic growth is the explanatory variable and government financial intervention is the threshold variable, the values are 1.1145 and 1.5295, respectively. When energy transformation is the explanatory variable and government financial intervention is the threshold variable, the values are 1.028 and 1.9285, respectively. When environmental improvement is the explanatory variable and government financial intervention is the threshold variable, the values are 1.1145 and 1.5309, respectively. When the significance level is 5%, the LR estimate corresponding to the threshold value of government financial intervention is significantly smaller than the critical value. Therefore, the above threshold estimate value is valid; that is, it passes the confidence interval test.

(5)

Regression analysis of the threshold for carbon neutrality in the economic growth, energy transformation, and environmental improvement under government financial intervention.

On the basis of the above work, this article uses Stata16.0 software for model regression, and

Table 9. Estimation results of threshold regression model parameters.

Explaining Variable	(1)	(2)	(3)
	$\mathit{E}\mathit{G}$	$\mathit{E}\mathit{T}$	$\mathit{G}\mathit{D}$
$EG\cdot I(GOV\le 1.1145)$	−0.14 ***
	(0.041)
$EG\cdot I(1.1145<GOV\le 1.5295)$	−0.024 *
	(0.032)
$EG\cdot I(GOV>1.5295)$	−0.016 ***
	(0.040)
$ET\cdot I(GOV\le 1.028)$		−0.043 *
		(0.038)
$ET\cdot I(1.028<GOV\le 1.9085)$		−0.149 ***
		(0.031)
$ET\cdot I(GOV>1.9085)$		−0.322 ***
		(0.063)
$EI\cdot I(GOV\le 1.1145)$			−0.133 **
			(0.044)
$EI\cdot I(1.1145<GOV\le 1.5309)$			−0.11
			(0.035)
$EI\cdot I(GOV>1.5309)$			−0.014 ***
			(0.046)
$OD$	0.04 **	0.03 ***	0.004 **
	(0.02)	(0.02)	(0.002)
$IS$	−0.03	0.015 *	−0.006
	(0.12)	(0.012)	(0.012)
$UL$	0.127 ***	0.109 ***	0.127 ***
	(0.029)	(0.026)	(0.026)
$TP$	−0.01	−0.001	−0.001
	(0.113)	(0.11)	(0.0117)
$IL$	0.006	−0.019 *	−0.0101
	(0.15)	(0.0136)	(0.0143)
$GOV$	−0.1092 ***	−0.0574 *	−0.1356 ***
	(0.0413)	(0.044)	(0.0445)
Constant	0.861 ***	1.0564 ***	1.0564 ***
	(0.189)	(0.1402)	(0.1402)
$R^2$	0.5585	0.5678	0.5678

Notes: ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively; the standard error is in parentheses.

reports the impact of government financial intervention on the carbon emission of EG, ET, and EI in Jiangsu Province. Within different intervals, regardless of the intensity of government fiscal intervention, its impact on carbon reduction varies in Jiangsu Province. Moreover, compared with the results in Table 3, the model has a higher goodness of fit ($R^2$), indicating the explanatory power of the panel threshold model is stronger, as shown in Table 9.

(1)

Analysis of threshold effect on economic growth.

According to the threshold effect model in column (1) of Table 9, it can be seen that Jiangsu provincial government fiscal intervention exhibits a significant threshold effect on the relationship between economic growth and carbon emission levels. The specific analysis is as follows. When the level of government fiscal intervention is less than 1.1145, its suppression of carbon emissions is most significant, with an impact coefficient of −0.14. As the level of fiscal intervention increases, its inhibitory effect on carbon emissions gradually weakens. When the government fiscal intervention level is within the range of 1.1145 to 1.5295, the impact coefficient drops to −0.024 and is significant at the 10% level. When the level of fiscal intervention exceeds 1.5295, the impact coefficient decreases to −0.016 but remains significant at the 1% level. In summary, the dual-threshold effect of fiscal intervention by the Jiangsu provincial government reveals a nonlinear relationship between economic growth and carbon emission reduction. With the increase in government fiscal intervention, the inhibitory effect of economic growth on carbon emissions gradually decreases, showing a diminishing marginal effect. Moderate fiscal support can effectively promote economic growth and curb carbon emissions, maximizing the effectiveness of fiscal intervention. This discovery provides an important theoretical basis for policy-making, emphasizing the need to pay attention to the appropriateness of intervention intensity when implementing fiscal policies.

(2)

Analysis of the threshold effect on energy transformation.

According to the threshold effect model in column (2) of Table 9, it can be seen that the fiscal intervention of the Jiangsu provincial government exhibits a significant threshold effect on the relationship between energy transformation and carbon emission levels. The specific situation is as follows. When the level of government fiscal intervention is less than 1.028, the energy transformation is associated with a statistically significant reduction in carbon emissions (at the 10% significance level). The level of government fiscal intervention has a negative effect within the range of 1.028 to 1.9085 and is significant at the 1% level, with a more prominent effect than when it is less than the 1.028 range. After exceeding 1.9085, it still exhibits a negative effect and is significant at the 1% level. With the increase in fiscal intervention, the government can effectively promote energy transformation through various means, thereby achieving carbon reduction targets. Strict energy transformation policies, combined with financial support, can encourage enterprises to adopt more proactive emission reduction measures and improve overall energy utilization efficiency.

(3)

Analysis of the threshold effect on environmental improvement.

According to the threshold effect model in column (3) of Table 9, it can be seen that the financial intervention of the Jiangsu provincial government exhibits a significant threshold effect on the relationship between environmental improvement and carbon emission levels. The specific situation is as follows. When the government fiscal intervention is less than 1.1145, the impact of environmental improvement on carbon emission levels is negative, with an impact coefficient of −0.133 and significance at the 5% level. The level of government fiscal intervention ranges from 1.1145 to 1.5309, showing a negative but not significant effect. After the level of fiscal intervention exceeds 1.5309, it has a negative effect and is significant at the 1% level, with an impact coefficient of −0.014. However, its inhibitory effect is significantly weakened compared to the first stage, which also reflects the diminishing marginal effect of government fiscal intervention. When implementing environmental governance policies, the strength and effectiveness of fiscal support are important factors in promoting carbon reduction. Therefore, the government should pay attention to the appropriateness of intervention intensity when formulating fiscal policies to ensure the efficient utilization of resources and the effective achievement of environmental governance goals.

4.3. Internal Mechanism Analysis

The above research has well confirmed the threshold effect and nonlinear relationship of government fiscal intervention on carbon emission levels in Jiangsu EG, ET, and EI. Figure 7, Figure 8 and Figure 9 illustrates the intrinsic mechanism of the nonlinear relationship between EG, ET, and EI and carbon emission reduction under the threshold effect of Jiangsu provincial government fiscal intervention.

As shown in Figure 7, with the increase in government fiscal intervention, economic growth effectively reduces carbon emissions and shows a marginal decreasing trend. This phenomenon is a regulatory effect of government fiscal intervention in this article. In the first range, moderate government fiscal intervention effectively promoted a positive interaction between economic growth and carbon emission mitigation. Economic growth significantly reduced carbon emission levels during this stage, demonstrating its positive contribution to carbon emission mitigation. This phenomenon indicates that under the reasonable guidance of fiscal policy, economic growth and environmental improvement can achieve coordinated development, reflecting the positive regulatory effect of government fiscal intervention. After entering the second interval, although economic growth still has a promoting effect on carbon reduction, the slope of this effect slows down significantly, indicating that the marginal effect of economic growth on carbon reduction is beginning to weaken. This may be because with further economic development, some traditional, high-energy-consuming industries still occupy a certain proportion, and the popularization and application of clean energy and green technology have not yet reached an ideal state, leading to a gradual weakening of the driving force of economic growth on carbon reduction. In the third interval, the inhibitory effect of government fiscal intervention on carbon emissions decreases, and the marginal effect weakens again. As the level of intervention increases, the marginal benefits of government intervention gradually decrease, which may be due to distortions in market mechanisms, the effectiveness of policy public transportation, and other reasons. In summary, government fiscal intervention has shown a significant role in moderating the relationship between economic growth and carbon emission levels. In different intervals, government fiscal intervention has had varying impacts on carbon emission mitigation by affecting the marginal effects of economic growth and the implementation of environmental protection measures. Therefore, the government needs to accurately grasp the intensity and direction of fiscal intervention, dynamically adjust fiscal policies in order to achieve coordinated development of economic growth and environmental improvement.

Figure 8 shows the sustained and significant negative impact of energy transformation on carbon emission levels, with a marginal increasing trend. This phenomenon is a regulatory effect of government fiscal intervention in this article. With the gradual increase in fiscal intervention, the absolute value of the slope shown in the graph (i.e., the quantitative indicator of the impact of energy transition on carbon emissions) shows an increasing trend, which clearly reveals that the effect of energy transformation on promoting carbon reduction is becoming increasingly significant. The steepness of the slope not only intuitively reflects the strong driving force of energy transformation itself in reducing carbon emissions but also deeply reflects the key role of government fiscal intervention in accelerating the energy transformation process and amplifying its emission reduction effect. In summary, the sustained negative impact of energy transformation on carbon emission levels shown in Figure 8 as well as the increasing absolute slope with increasing fiscal intervention fully demonstrate the powerful regulatory effect of government fiscal intervention in promoting energy transformation and carbon reduction. This not only reflects the important role of government policies in guiding energy structure optimization and promoting green and low-carbon development but also provides strong empirical support for formulating more scientific and effective energy and environmental policies in the future.

Figure 9 reveals the inhibitory effect of environmental improvement on carbon emissions under different intensities of government fiscal intervention, which shows a marginal decreasing trend. This phenomenon is a regulatory effect of government fiscal intervention in this article. In the first range, moderate government fiscal intervention provides necessary financial support and policy guidance for environmental improvement, enabling effective implementation of environmental improvement measures and significantly suppressing carbon emission levels, successfully reducing carbon emissions. At this stage, the regulatory effect of government fiscal intervention is positive, effectively promoting the coordinated development of environmental improvement and carbon reduction through optimizing resource allocation, strengthening environmental supervision, and other means. However, when entering the second interval, similar to economic growth, the effect of environmental improvement on carbon emission levels significantly decreases, and the marginal effect of policy tools begins to decrease. The government may face problems such as decreasing policy effectiveness and increasing fiscal budget pressure. The inhibitory effect of government fiscal intervention on carbon emissions in the third interval continues to decrease, which may be due to an insufficient innovation ability of enterprises and the saturation of the technological level, leading to a weakening of the impact of environmental improvement on carbon emissions. The government needs to continuously improve policy tools and enhance the efficiency of policy implementation to ensure the effectiveness of environmental improvement. In summary, the different effects of environmental improvement on carbon emission levels under different government financial interventions shown in Figure 9 not only reveal the moderating effect of government financial interventions on the relationship between environmental improvement and carbon reduction. It is necessary to strengthen the supervision and evaluation mechanism and timely discover and solve problems in policy implementation to achieve the coordinated development goal of environmental improvement and carbon reduction.

5. Discussion

Based on the above model construction and empirical result analysis, this paper draws the following findings.

(1)

The economic development of Jiangsu Province has a U-shaped relationship with carbon emissions, which means that during the process of economic development, carbon emissions will first decrease and then increase with economic growth. There is an inverted U-shaped relationship between energy transformation and carbon emission levels. In the process of energy transformation, carbon emissions will first increase and then decrease with the energy transformation. Environmental improvement in Jiangsu has always had a negative effect on carbon emissions, reflecting the role of the environmental improvement system in promoting carbon emission reduction. Environmental improvement suppresses carbon emissions by formulating and implementing policies such as environmental standards, emission limits, and economic instruments, prompting enterprises and individuals to adopt low-carbon technologies and management measures. Based on the principle of the environmental Kuznets curve, Fan et al. [21] contend that an empirical analysis of panel data reveals a suppressive influence of green finance development on carbon emissions. The development of an innovative economy can effectively reduce carbon dioxide emissions and enhance carbon productivity, mainly through upgrading industrial structure, stimulating social public awareness of environmental protection [59], promoting technological diversification, and improving the efficiency of energy use [60] and other strategies to realize the low-carbon transformation of the production and lifestyle, and then improve the regional carbon emission reduction energy transformation through energy transformation policy, promote technological innovation, and improve environmental regulation so as to realize carbon emission reduction [22]. Overall, the widespread application of energy innovations reduces the demand for fossil fuels and provides an opportunity to accelerate the clean energy transformation globally. Global energy demand growth in 2023 is stronger than in 2022, and the rapid advancement of the clean energy transformation plays a role in controlling carbon dioxide emissions. EG, ET, and EI are mutually reinforcing in achieving carbon neutrality. The interactions within the three systems are complex, and their optimal state is manifested in a coordinated relationship among the three subsystems [61]. Energy transformation, environmental improvement, and the sustainability of economic growth are interdependent and indispensable, while the coordination among the three promotes the stability and sustainable development of the whole system, thus effectively reducing carbon emissions [62].

(2)

This article conducts an in-depth analysis of provincial government financial intervention as a threshold-moderating variable. The research results show that the degree of government financial intervention has a significant moderating effect on the effect of energy transformation on suppressing carbon emissions. Specifically, as the degree of government fiscal intervention increases, the inhibitory effect of energy transformation on carbon emissions gradually strengthens. Dong et al. [63] studied the carbon emission reduction effect of the low-carbon pilot policy on high-energy-consuming industries and found that the policy can effectively reduce the carbon emissions of high-energy-consuming industries and achieve carbon emission reduction by promoting industrial upgrading and optimizing the energy structure. The new energy demonstration city policy is able to reduce carbon emissions by increasing the intensity of government environmental regulation, mainly through environmental regulation to reduce carbon emissions from regional and manufacturing industries [64]. In order to develop the new energy industry, local governments incentivize technological innovation, actively guide the direction of investment, and enhance the endogenous motivation of technological innovation of enterprises [65]. Energy transformation technology innovation activities are longer, riskier, and more difficult than traditional technology innovation activities, and effective policies can guide the flow of funds to new energy technology innovation [66].

(3)

Using Jiangsu provincial government financial intervention as a threshold adjustment variable, this study explores the impact of government financial intervention on the relationship between economic growth, environmental improvement, and carbon emissions. The research results show that government fiscal intervention plays a key regulatory role. This study indicates that when the degree of government fiscal intervention is low, economic growth and environmental improvement have a negative impact on carbon emissions, that is, to suppress carbon emissions. However, when the level of government fiscal intervention surpasses a certain threshold, both economic growth and environmental improvement paradoxically exhibit a positive correlation with carbon emissions, effectively promoting them. This means that moderate government financial intervention is favorable to economic growth and environmental improvement to promote carbon emission reduction, while excessive government financial intervention may inhibit carbon emission reduction. A moderate environmental target responsibility system can improve environmental quality, promote technological innovation, and enhance social welfare, which in turn is conducive to carbon emission reduction. However, if the government intervenes excessively, it may inhibit economic growth and technological innovation, which ultimately affects the effect of carbon emission reduction [67]. The effectiveness of China’s carbon market in reducing carbon emissions was found to diminish as the intensity of government administrative intervention increased. This suggests that excessive intervention may distort the market mechanism, inhibit the enthusiasm of enterprises to reduce emissions on their own, and ultimately affect the effect of carbon emission reduction [42]. Government financial intervention influences both the economy and environment of a region, with a higher level of intervention leading to a greater degree of involvement in both these domains. The government, as the main body of regulation and resource allocation, is an important guiding force for realizing green development and carbon neutrality [68]. Local government financial intervention has a positive effect on industrial transformation and upgrading and environmental quality improvement, and the stronger the intervention is, the more obvious the effect on environmental quality improvement is [69]. In the short term, governments at all levels, as policy decision-makers and implementers, are limited in the growth of fiscal revenues due to the implementation of tax cuts and fee reductions, which restricts the effective play of the government. In the long run, tax and fee reductions will encourage enterprise innovation, improve productivity, and ensure revenue growth, thus realizing a “win-win” situation for both the government and enterprises [70]. For regions with a low level of government fiscal intervention, local governments prefer gradual over mandatory environmental regulations, and such guided interventions may rely on external mechanisms to drive green transformation. In regions with high government fiscal intervention, the impact of environmental information disclosure quality on enterprise green transformation is weakened, suggesting a threshold effect. Nevertheless, environmental information disclosure remains crucial for green transformation. This indicates that there is a certain threshold effect of government fiscal intervention [71]. Government financial intervention can significantly affect carbon emission efficiency, and the specific effect depends on whether the government financial intervention is a “crowding out effect” or a “forcing effect” [72].

(4)

This study reveals that while factors such as the level of openness to external entities, industrial structure upgrading, and urbanization contribute to economic growth, they also tend to be accompanied by increased energy consumption, which hinders carbon emission reduction efforts. On the contrary, technological progress, the enhancement of the informatization level, and the strengthening of government financial intervention can effectively reduce the intensity of energy consumption, which in turn reduces carbon emissions and promotes carbon emission reduction. Some scholars have proposed that urbanization will accelerate the growth of carbon emissions mainly by affecting the economy [73]. In the research of BRICS countries, urbanization and carbon emissions show a relationship of mutual constraints and interactions [74]. Urbanization causes an increase in energy consumption and environmental pressure, inhibiting carbon emission reduction, and the process of urbanization must be subject to reasonable disruption [75]. Industrial structure transformation can promote structural rationalization, strengthen inter-industry coordination and cooperation, optimize resource allocation, promote industrial transformation and structural upgrading, improve overall production efficiency, and effectively reduce carbon emissions [76]. Some scholars believe that the rationalization of the industrial structure is correlated with carbon emissions, while industrial structure upgrading is conducive to reducing carbon emissions [77,78]. Technological progress promotes the improvement of dynamic carbon emission efficiency, and technological progress has a significant effect on its convergence [79]. Wen et al. [80] found that technological innovation is an important mechanism for green finance to exert emission reduction effects. To effectively control carbon emissions, it is crucial to balance economic development with environmental protection, optimize economic structures, foster green and low-carbon tech innovation, and strengthen government–market synergies.

6. Conclusions and Policy Recommendations

6.1. Main Conclusions

This article is based on the DPSIRM theoretical framework model. First, a comprehensive evaluation index system for the coordinated development of EG, ET, and EI is constructed. The improved entropy method is used to measure the comprehensive level of the three systems in sequence. We calculated the carbon emissions in Jiangsu Province from 2010 to 2020 using the carbon emission factor method and analyzed the trend of carbon emissions in Jiangsu over the past decade. Using government fiscal intervention as a threshold variable and constructing a threshold effect model, we identified the evolutionary characteristics of economic growth, energy transformation, and environmental improvement on carbon emission levels under local government fiscal intervention. The results show the following:

(1)

The energy transformation in Jiangsu Province has an inverted “U”-shaped relationship with carbon emissions, while economic development has a “U”-shaped relationship with carbon emissions. Environmental improvement has a significant negative impact on carbon emissions. Within the research interval, carbon emissions will first increase and then decrease during the energy transformation process. This discovery highlights the importance of policy support for technological innovation and infrastructure development, especially in the early stages of energy transformation. Although there may be short-term increases in carbon emissions, effective policy interventions can accelerate the transition to a mature low-carbon energy system, ultimately achieving a significant reduction in carbon emissions. In the process of economic development, carbon emissions will first decrease and then increase with economic growth. At the beginning, economic development may lead to a reduction in carbon emissions due to technological progress and industrial structure optimization. With sustained economic growth and increasing population, energy demand may rise again, highlighting the urgent need for proactive environmental policies and strategic energy transformation to decouple economic growth from carbon emissions exceeding a certain threshold. Contrary to the nonlinear relationship observed in economic growth and energy transformation, environmental improvement in Jiangsu Province has a sustained negative effect on carbon emissions, further confirming the effectiveness of strict environmental regulations, emission limits, and economic tools in promoting sustainable practices and technology adoption. The sustained inhibitory effect of environmental governance not only demonstrates the crucial role of effectively implemented environmental policies in achieving sustained carbon emission reduction but also emphasizes the crucial importance of the sustainability and stability of policy interventions in ensuring emission reduction effectiveness throughout the entire transformation process. This study provides empirical evidence for this collaborative relationship in the specific context of Jiangsu Province. The U-shaped curve of economic growth and the inverted U-shaped curve of energy transformation highlight the importance of strategic policies to maximize positive impacts and minimize negative consequences during the transition period. The sustained inhibitory impact of environmental improvement highlights the need for effective policy intervention throughout the entire process.

(2)

This article found that local government fiscal intervention has a significant effect on carbon reduction changes, which is called the regulatory effect of fiscal intervention. Specifically, with the increasing degree of government fiscal intervention, energy transformation, economic growth, and environmental improvement have all shown positive impacts on reducing carbon emissions, and the impacts have shown a differentiated marginal effect trend. In terms of energy transition, its contribution to carbon reduction shows an increasing marginal effect, which means that with the increase in fiscal support and the optimization and upgrading of energy structure not only continue to advance, but also its emission reduction effect becomes more significant. This reflects the unique advantages and high efficiency of fiscal policy in accelerating energy technology innovation and promoting the development of new energy. In contrast, the marginal effect of carbon reduction in the fields of economic growth and environmental improvement shows a decreasing trend. Although economic growth can indirectly promote carbon reduction, with the expansion of economic scale, its inherent carbon emission pressure may gradually increase, requiring more refined policy regulation to balance the relationship between development and emission reduction. In the field of environmental improvement, the increasing difficulty of governance and rising marginal costs may lead to a gradual decrease in emission reduction efficiency, requiring the government to continuously innovate governance models and increase investment. The increase in the marginal rate of return in this study indicates that fiscal intervention measures have been successful in creating a supportive environment for energy transformation.

(3)

This study deeply analyzed the economic and social development factors, such as the degree of opening up to the outside world, upgrading of industrial structure, and improvement of the urbanization level, as well as the impact mechanism of policy measures such as technological progress, the improvement of information technology level, and government financial intervention on carbon emissions, revealing the complex role of these factors in the process of carbon reduction. Specifically, the increase in openness to the outside world, the upgrading of industrial structures to a higher level, and the rapid improvement of urbanization have to some extent promoted economic and social development, but at the same time, accompanied by the growth of energy consumption, they have also had a restraining effect on carbon emission mitigation. This discovery indicates that while enjoying the economic benefits brought by globalization dividends, industrial structure optimization, and accelerated urbanization, it is necessary to pay close attention to the potential environmental pressures it may bring, especially the increase in carbon emissions, in order to avoid falling into the development trap of “pollution first, treatment later”. However, research has also found that technological progress, improvement in information technology, and increased government financial intervention are important ways to effectively reduce energy consumption intensity, decrease carbon emissions, and promote carbon reduction. Technological progress has reduced carbon emissions from the source by improving energy efficiency and developing clean energy and low-carbon technologies. The improvement of the informatization level has reduced energy consumption and carbon emission intensity through optimizing resource allocation, improving management efficiency, and promoting information sharing. The strengthening of government fiscal intervention, through means such as financial support, policy guidance, and market supervision, has promoted the development of green and low-carbon industries, suppressed the disorderly expansion of high-energy-consuming and high-emission industries, and thus achieved effective control of carbon emissions.

6.2. Policy Recommendations

This study provides new ideas for the implementation of government fiscal intervention policies by analyzing in depth the regulatory effects of government fiscal intervention on carbon reduction and offers a new perspective for high-quality economic development, optimizing energy transformation paths, achieving environmental green development and carbon neutrality. To this end, this study derives the following policy recommendations:

(1)

We should stimulate industrial transformation and promote the ecological transformation of conventional industries; we should improve green finance, promote the innovative development of green financial products, and provide more convenient and low-cost financial support for the green industry. We should encourage green consumption, advocate green lifestyles and develop the market for sustainable products and services. We should vigorously develop renewable energy and formulate development goals and support policies. We should optimize energy utilization efficiency; intensify the research, development, and deployment of energy-saving technologies; elevate energy productivity; and minimize energy consumption waste. We should build an intelligent energy system and promote the construction of the energy Internet. We should strengthen the management of environmental pollution, formulate environmental protection standards, and increase environmental enforcement. We should protect the ecological environment, promote ecological restoration, maintain the balance of ecosystems, and improve the carbon sink capacity of ecosystems. We should encourage green scientific and technological innovation and promote green technological innovation in enterprises.

(2)

We should optimize the design of fiscal policy and avoid excessive intervention. The government should formulate scientific and reasonable fiscal policies, avoid over-reliance on financial subsidies, focus on playing the pivotal role of market mechanisms in allocating resources, and guide enterprises to innovate independently to realize sustainable carbon emission reduction. We should strengthen the precise input of financial funds. The government should focus financial funds on key areas and weak links, such as promoting the application of clean energy and improving the carbon emission trading system so as to enhance the effectiveness of financial fund utilization. We should strengthen the synergy between government guidance and market mechanisms. The government ought to enhance its guidance and oversight of the energy transformation process, and it should utilize market mechanisms effectively, encourage enterprises to actively participate in energy transformation, and form a good situation of government guidance, with a market-driven, enterprising main body and social participation.

(3)

We should set scientific and reasonable carbon emission targets and formulate corresponding carbon emission control programs. We should develop a robust carbon emission trading market and leverage market mechanisms to incentivize enterprises to cut down on carbon emissions. We should implement carbon tax policy, levy carbon tax on high-carbon-emission industries and promote carbon emission reductions among businesses. The government fiscal intervention policy needs to be combined with the market mechanism. Carbon neutrality is a complex systematic project that requires the government to set clear carbon neutrality targets and timetables, and in this regard, the government needs to encourage enterprises and individuals to actively participate in carbon emission reduction through policy guidance so as to form a situation of joint efforts by the whole society. Environmental pollution and climate change caused by carbon emissions are typical negative externalities that cannot be effectively addressed by market mechanisms. The government needs to internalize the cost of carbon emissions and guide enterprises to reduce carbon emissions by formulating carbon emission standards, levying carbon taxes, and implementing carbon emission trading and other policies.

(4)

Enhancing openness will facilitate the introduction of advanced emission reduction technologies, thereby boosting the overall efficiency of carbon reduction in industries. Additionally, the tertiary industry boasts relatively low carbon emissions, and its development is beneficial for decreasing carbon emission intensity. Strengthening the construction of urban infrastructure can effectively reduce carbon emissions. The improvement of the urban management level can effectively reduce energy waste and improve carbon emission reduction efficiency. Advances in new energy technology can provide clean, low-carbon energy alternatives; improving the level of informatization is conducive to information sharing, promoting coordination and cooperation among all parties, and improving the efficiency of carbon emission reduction. Informatization technology can realize accurate energy management, effectively reduce energy waste, and improve carbon emission reduction efficiency.