The Impact of Heterogeneous Environmental Regulation Tools on Economic Growth: Can Environmental Protection and Economic Growth Be Win-Win?

Abstract

This paper explores the relationship between economic growth and environmental regulation using panel data for 30 provinces in China from 2009–2021 using the fixed effects model and the threshold model. First, the baseline regression results show that market-based environmental regulation promotes economic growth. However, command-and-control environmental regulation in China can not promote China’s economic growth. Second, further research has shown that environmental regulation’s role in promoting the economy is constrained by the intensity of environmental regulation. Market-based environmental regulation is only able to promote economic growth when the intensity is low. If the intensity of market-based environmental regulation is too high, market-based environmental regulation, on the contrary, can not promote economic growth. Third, The impact of market-based environmental regulation on economic growth also is found to be constrained by the level of economic development: market-based environmental regulation does not promote economic growth when the economy is less developed. It is only when the economy has reached a high level that market-based environmental regulation will contribute to economic growth. Finally, this paper finds that financial development and market-based environmental regulation can synergize to promote economic growth.

1. Introduction

China’s economy has had a four-decade fast development boom since its reform and opening-up. During this process, China has prioritized economic construction, making economic growth its top priority. Simultaneously, GDP has become the key indicator for assessing local government performance, with the promotion of local officials directly tied to the local economic figures. Local governments have excessively pursued economic growth, long ignoring the balance between economic growth and environmental protection. In the process of economic growth, uncontrolled consumption of resources has exacerbated environmental damage. Economic growth is determined by a variety of factors, but China’s environmental pollution problem has greatly constrained China’s economic growth. In 2007, a joint study from the World Bank and the Chinese government, “Environmental Pollution Losses in China”, showed that air and water pollution cost the Chinese economy around USD 100 billion per year. Environmental pollution has gravely impacted the health of the Chinese public, tarnished the government’s image, and constrained the quality of China’s economic development. The era of China’s most rapid economic expansion has also been marked by its most severe environmental pollution challenges. Serious environmental problems pose a huge challenge to the development of China’s economy. China is under tremendous pressure from the international community to lower its energy consumption and carbon dioxide emissions, given that it is currently the world’s greatest emitter of carbon dioxide and user of energy. China’s development has long relied on high energy consumption and emissions, adopting a careless approach that excessively sacrifices environmental quality, which is unsustainable [1]. As the level of economic development in China increases, the Chinese people’s aspirations for the quality of their living environment are getting higher. To improve environmental conditions, it is imperative to rigorously control environmental pollution. This is crucial not only for reducing pollution but also for fundamentally transforming China’s economic development method. Restrictions on products made with environmentally harmful technologies are now increasingly being imposed in regions such as the European Union. As a major trading nation, China must also take into account the environmental expectations of importing countries. The implementation of environmental regulations is of crucial importance to maintain China’s export competitiveness.

There is no doubt that environmental regulation is the fundamental institutional arrangement for solving the problem of environmental pollution [2]. As a public good, the problem of externalities predisposes the environment to abuse. Environmental protection may not be achieved simply by the voluntary acts of enterprises and therefore requires the involvement of the government [3]. By defining the externalities of natural resource assets and pollution emissions, environmental regulation makes up for the inadequacies of the market mechanism, raises the cost of environmental resources for emission-emitting enterprises, and motivates them to reduce pollution. Some viewpoints believe that energy conservation and emission reduction affect economic development and that economic development can only be achieved at the expense of the environment, and they argue that environmental regulations imposed by the government inevitably place a burden on businesses, perhaps leading to a drop in production, weakening their competitiveness and causing them to exit the industry or relocate to areas with weaker environmental regulations [4]. Under this logic, protecting the environment inevitably leads to a decline in economic growth rates. Many Chinese local governments have pitted environmental protection against economic growth, pursuing economic growth at the expense of environmental protection, meaning many environmental regulation policies have not been effectively implemented. If only the local effects of environmental protection are taken into account, it can be argued that environmental protection will hinder economic development, thus leading to a dilemma: should we protect the environment or develop the economy first? However, this view is not consistent with classical economic growth theory. Neoclassical economic growth theory tells us that the ultimate driver of economic growth is productivity growth, yet environmental pollution can reduce labor productivity levels [5]. Thus, in the long-term, protecting the environment not only does not impede economic growth but also promotes long-term economic growth. In addition, the emergence of the Porter Hypothesis offers a more optimistic perspective on the relationship between environmental regulation and economic performance. Proposed by Porter [6], this hypothesis posits that well-designed environmental regulations can stimulate innovation, enhance competitiveness, and lead to greater economic benefits.

So, what is the relationship between environmental regulation and economic growth in China currently? This paper mainly focuses on the effects of different types of environmental regulatory tools on economic growth in China and only chooses and mathematically models the relationship between environmental regulation and economic growth and disregards other factors determining economic growth in China. At present, do environmental protection and economic growth inherently conflict, or can they coexist in a win–win scenario? Such an analysis is not only of theoretical importance but also carries significant practical implications for China’s pursuit of sustainable, green development. It will help resolve the ongoing dilemma of prioritizing either environmental protection or economic development first. This paper explores the impact of different types of environmental regulations on economic growth using panel data for 30 provinces in China from 2009 to 2021.

The following are the primary aspects to which this paper contributes: Firstly, this paper provides macro evidence from China for the debate on whether environmental regulation and economic growth are in “conflict” or “win-win”. This paper constructs linear and non-linear models to clarify comprehensively the relationship and constraints between environmental regulation and economic growth. Secondly, this paper intends to explore how environmental regulations influence economic growth from two distinct approaches: market-based and command-and-control. China’s environmental regulation strategy provides an ideal case to study the different impacts of these two types of regulation on economic growth. This paper provides the basis for the government to designate and implement environmental regulatory policies at this stage.

2. Literature Review and Theoretical Mechanisms

As the importance of environmental resources has been increasingly recognized, how to reconcile the relationship between the environment and economic growth has been the focus of attention of economists and scholars in various countries. There is a lot of literature that has studied the impact of environmental regulation on economic growth but has come to two views with opposite conclusions.

The traditional neoclassical school of environmental economics generally believes that the costs are such that environmental regulation creates a heavy economic burden on the regulated firms and this could affect total output and have a negative impact on the economic growth [7,8]. The focus of attention in this research is the cost of environmental regulation, including (1) the direct costs associated with pollution abatement; and (2) the opportunity costs associated with firms reducing investment in other innovative projects in order to increase investment in pollution abatement. In particular, as a result of the government’s environmental regulation policy being put into effect, the enterprise’s operating costs will rise in order to comply with environmental protection standards in their production and operation processes. This compliance can be achieved by increasing investment in pollution control measures, reducing the intensity of pollutant emissions, or paying pollution charges. This will result in higher operational expenses for the business, i.e., the environmental regulation of the production activities of the enterprise will have a “cost effect” [9]. The increase in production costs will cause enterprises to reallocate resources and will lead to the shift of labor and capital from productive activities to environmental protection and pollution control. This will, to some degree, impede business output and technical advancement. It also does not help businesses become more productive, which impedes economic development.

In response to the cost–effect view, Porter [6] put forward a different idea, the Porter hypothesis, arguing that rationally designed regulatory policies can stimulate firms’ innovative behavior, improve productivity and reduce costs. This thereby offsets the additional costs of environmental regulation and the innovation inputs themselves, i.e., environmental regulation produces an “innovation compensation effect” [10,11]. According to the Porter hypothesis, the conclusion drawn by focusing only on the cost effect is incorrect and insufficient to explain the impact of environmental regulation on economic growth. If considering the innovation compensation effect of environmental regulation, it is something that will ultimately promote economic growth. Porter’s hypothesis has been supported by many scholars since it was proposed [12,13,14,15]. However, some scholars have also questioned the Porter hypothesis [16]. Palmer et al. [17] argue that the effect of innovation compensation is minuscule relative to the cost effect and innovation offsetting costs is only a special case. There is ongoing debate over how environmental regulations affect economic growth.

The analyses above suggest that environmental regulation has two primary effects: the cost effect and the innovation compensation effect. Therefore, if the cost effect outweighs the innovation compensation effect, environmental regulation is likely to inhibit economic growth. On the flip side, if the benefits of innovation resulting from environmental regulations exceed the associated costs, then such regulations are likely to boost economic growth. The net effect of environmental regulations on economic growth thus hinges on the balance between these advantageous and adverse impacts.

Market-based and command-and-control environmental regulations are the two categories into which some of the literature divides China’s environmental regulations [18,19,20,21]. Market-based environmental regulation is an economic incentive provided by the government to guide companies in pursuing their own interests to achieve pollution control goals [18]. In China, market-based environmental regulations include (1) fees and taxes related to environmental pollution. When companies or other producers and operators release taxable pollutants into the environment, they must record the release and comply with the administration and collection processes by paying an environmental protection tax or a pollution discharge fee; (2) Emission licenses. Enterprises can trade licenses with other emission units according to their own emission situation and under the premise of ensuring their own emission needs [20]. In addition, China’s environmental regulations have been command-and-control in large part. This method is primarily characterized by its coercive nature, relying on administrative directives [22]. Polluting enterprises are ordered and supervised by government departments to control pollution emissions within the required limits by reducing production or improving production technology.

Market-based environmental regulation offers greater flexibility to firms, which can significantly foster technological innovation. This flexibility allows firms to choose the most cost-effective and technologically advanced methods of compliance, encouraging creative solutions [23]. On the other hand, command-and-control regulations, which are typically more rigid and prescriptive, mandate specific actions that firms must take. This can lead to higher compliance costs, as firms may need to invest in specific technologies or processes regardless of their current operations or efficiency. Thus, market-based approaches are more conducive to innovation efficiency, while command-and-control approaches might increase financial burdens on firms due to their inflexible requirements and be not conducive to innovation. Thus this paper proposes the following hypothesis:

H1. 

Market-based environmental regulation can promote economic growth in China but command-and-control environmental regulation may not promote economic growth in China.

Exploring the effects of environmental regulation requires a detailed analysis that not only accounts for the intensity of the environmental regulations but also considers broader macroeconomic conditions and other pertinent factors [24,25,26]. As the intensity of environmental regulation changes, the magnitude of the cost effect and the innovation compensation effect also changes, thus changing the overall impact effect on economic growth [27,28,29,30]. Increases in regulatory intensity may impose additional costs on firms, thereby discouraging investment and reducing economic output. If environmental regulations are too stringent and bring about a cost effect that is too large, then the innovation effect will not be able to compensate for the cost effect, and economic growth will be hampered [31]. In regions with low levels of economic development, firms tend to be more underfunded, and financial markets may be underdeveloped, making it difficult for firms to obtain financial support [32]. Even a low intensity of environmental regulation crowds out funds for routine economic development, thus seriously hampering the normal production and operation of firms. Thus this paper proposes the following hypothesis:

H2. 

The impact of environmental regulation on economic growth could be non-linear, changing with different levels of regulatory intensity.

H3. 

The effect of environmental regulation on economic growth might also be influenced by the level of economic development.

Ignoring the issue of corporate financing constraints when enforcing environmental regulations to promote economic growth may not only fail to promote changes in production methods but also magnify the environmental risks faced by micro-firms and exacerbate economic volatility [33]. At the micro level, stringent environmental regulations can lead to higher environmental costs for polluting firms, with firms’ marginal costs exceeding their marginal benefits. In this scenario, when the cost of financing environmental remedies is high, firms may have smaller benefits from pollution remedies and thus refrain from active pollution remedies and reduce pollution only by reducing economic activity [34], which ultimately has a negative impact on economic output. Thus, this paper proposes the following hypothesis:

H4. 

Environmental regulation and financial development can synergize to promote economic growth.

The Figure 1 shows the flowchart of the research in this paper:

3. Empirical Model Setting and Data

3.1. Definition of Variables

(i)

Economic growth

Both overall GDP and GDP per capita can reflect economic growth. Referring to Barro [35] and Song et al. [27], this paper uses the GDP per capita growth rate to measure economic growth. This is because GDP per capita is a better indicator of the real economic development level change than overall GDP [27].

(ii)

The environmental regulation

Numerous approaches for quantifying environmental regulation are discussed in the literature. For instance, Song et al. [36] employed pollution emission intensity, defined as the ratio of pollutants to industrial output, to assess environmental regulation. Similarly, Tang et al. [37] utilized pollution control costs as an indicator of environmental regulation. Other researchers like Yang et al. [38], and Ji et al. [39] measure environmental regulation by industrial pollution control investment. Additionally, Yin et al. [40] used the number of environmental administrative penalties and the cumulative number of laws enacted across regions as metrics for environmental regulation. The existing literature has used different indicators to measure environmental regulation and has produced different results: some literature has found that environmental regulation promotes growth [41,42], while others have found that it does not promote, or even hinder, economic growth [18]. This paper intends to further explore whether different types of environmental regulation tools have different impacts on economic growth.

Referring to some of the literature’s practices, this paper categorizes environmental regulation into two types: market-based environmental regulation and command-and-control environmental regulation [18,19,20].

Referring to Pan et al. [19] and Guo and Yuan [43], this paper uses the ratio of pollution discharge fee to industrial added value to measure the strength of market-based environmental regulation tools. The symbol MBER stands for market-based environmental regulation.

$$\mathrm{MBER}=\frac{\mathrm{Total} \mathrm{Pollution} \mathrm{discharge} \mathrm{fee} \mathrm{payments}}{\mathrm{Industrial} \mathrm{added} \mathrm{value}}$$

Pollution charges are levied on air pollutants, water pollutants, and solid waste. Pollution charges are levied only on companies (mainly industrial companies), not on individual residents. Starting in 2018, China initiated a reform of its pollution discharge fee system and officially enacted an environmental protection law on 1 January 2018. Subsequently, an environmental protection tax took the place of the previous pollutant discharge fee. Consequently, in this paper, the data pertaining to pollution discharge fees from 2018 to 2021 have been substituted with figures from environmental tax collections. Whether it is called a pollution charge or an environmental tax, their purpose is to hold businesses accountable for the unfavorable externalities of their operations. So it will not affect the conclusions of this paper. Also, this paper ensures the robustness of the results by using only the data for the period 2009–2017 in the robustness test.

Referring to the practice of Liu et al. [18] and Xie et al. [20], this paper uses the ratio of the Industrial Pollution Control Investments to Industrial added value to measure the strength of cost-based environmental regulation. The symbol CACER stands for command-and-control environmental regulation. The data on the Industrial Pollution Control Investments are from the China Environmental Statistics Yearbook.

$$\mathrm{CACER}=\frac{\mathrm{Industrial} \mathrm{Pollution} \mathrm{Control} \mathrm{Investments}}{\mathrm{Industrial} \mathrm{added} \mathrm{value}}$$

(iii)

Control variable

This paper includes a set of provincial characteristic control variables in the regression model. Referring to the practice of Song et al. [27], Zhao et al. [29] and Zhang et al. [44], this set of variables includes openness, scientific and technological inputs, industrial structure, and physical capital investment:

Openness (OPEN): Since the economic reforms and opening-up policy initiated in 1978, China’s economy has experienced rapid growth, with a consistent increase in regional GDPs. The success of the opening-up policy has been profound, significantly contributing to the economic prosperity of various regions. This paper defines that OPEN = FDI/GDP.

Scientific and technological inputs (ST): Scientific and technological inputs can influence economic growth through innovative activities. The ratio of expenditure on research and development to GDP is used as a measurement indicator of Scientific and technological inputs.

Industrial structure (IS): The pattern of regional economic development is influenced by the industrial structure. China’s economic development is currently going through a significant upgrade transitional phase [27]. The evolution of the industrial structure will inevitably affect the economy. The ratio of industrial-added value to GDP is adopted for measurement in this paper.

Physical capital investment (PCI): Investment in physical capital plays a vital role in influencing the distribution of resources and production efficiency. Physical capital investment is denoted by the proportion of real fixed asset investment to GDP.

3.2. Empirical Model Setting

In order to examine the heterogeneous impact of environmental regulation on economic growth, this paper constructs the following equation:

Growth_it = α + β₁MBERit + β₂CON_it + d_Province + d_year + ε_it

(1)

Growth_it = α + β₁CACERit + β₂CON_it + d_Province + d_year + ε_it

(2)

By using the Hausman test, the results of the test suggest that fixed effect is more suitable than random effect. Thus, this paper uses a fixed effects model, where Growth_it is the economic growth of province i in year t and MBER_it represents market-based environmental regulation. CACER_it represents command-and-control environmental regulation.

The coefficient β₁ measuring the impact of environmental regulation on economic growth is therefore the central parameter of interest in this paper. In order to minimize the bias caused by omitted variables, this chapter additionally accounts for a set of control variables related to province characteristics in the baseline regression model, denoted by CON_it. The d_Province and d_year are individual fixed effects (i.e., province fixed effects) and time fixed effects, respectively. Finally, ε_it is the error term.

3.3. Statistical Description of the Data

In this paper, panel data covering 30 Chinese provinces between 2009 and 2021 are used. The sources of the data include the China Environmental Statistics Yearbook, the China Statistical Yearbook, the National Bureau of Statistics of China website, and the China Regional Statistical Yearbook.

Table 1. Statistical description of the data.

Variable	Symbol	Obs	Mean	Std. Dev.	Min	Max
Economic growth (%)	Growth	390	8.020	3.250	−3.600	18.800
Market-based Environmental regulation (%)	MBER	390	0.074	0.053	0.007	0.423
Command-and-control environmental regulation (%)	CACER	390	0.245	0.221	0.005	2.035
Industrial structure	IS	390	0.435	0.088	0.158	0.590
Openness (%)	OPEN	390	1.937	1.584	0.006	8.191
Scientific and technological inputs (%)	ST	390	2.692	0.371	1.396	3.288
Physical capital investment (%)	PCI	390	13.455	0.812	10.687	14.804

reports the statistical description of the data. Testing reveals that all of the variance expansion factors (VIFs) are less than 10, indicating that there is not a significant issue with multiple collinearity between the interpretative variables.

4. Empirical Results and Analysis

4.1. Baseline Regression Results

Based on the panel data of 30 provinces in China from 2009 to 2021, this paper first applies the fixed effects model to estimate Equations (1) and (2).

Table 2. Baseline regression results.

Variables	(1)	(2)	(3)	(4)
MBER	4.840 *	6.225 **
	(1.72)	(2.21)
CACER			−0.099	0.729
			(−0.18)	(1.28)
IS		8.207 **		7.725 **
		(2.42)		(2.27)
OPEN		0.348 ***		0.352 ***
		(3.48)		(3.50)
ST		−25.985		−46.155
		(−0.30)		(−0.54)
PCI		−0.818		−0.890
		(−1.24)		(−1.34)
Constant	10.324 ***	5.992 ***	110.946 ***	6.759 ***
	(24.65)	(3.52)	(341.68)	(4.08)
Observations	390	390	390	390
Fixed effects	YES	YES	YES	YES
R2	0.761	0.775	0.807	0.773

(t value in parentheses, *** p < 0.01, ** p < 0.05, * p < 0.1).

reports the regression results.

The first column reports the results of MBER (market-based environmental regulation) without a control variable. The second column reports the regression results of MBER with control variables. It can be seen that the regression coefficients for MBER are significantly positive in both the first and second columns. This shows that China’s MBER does contribute to China’s economic growth.

The third column reports the results of CACER (command-and-control environmental regulation) without a control variable. The fourth column reports the regression results of CACER with control variables. It can be seen that the regression coefficients for CACER are not significant in both the third and fourth columns. This shows that China’s CACER does not promote economic growth.

This result proves Hypothesis 1. MBER is more likely to be a driver of innovation than CACER regulation. The MBER acts as an economic incentive mechanism by creating an economic incentive for enterprises that emit pollutants to take the initiative in seeking innovative emission reduction solutions. Driven by economic interests, enterprises will more actively seek more efficient and environmentally friendly production methods. In contrast, CACER may make firms take a more fixed and standardized approach and technology, thus limiting firms’ pursuit of innovation. In addition, CACER has a higher cost effect due to its coercive nature and more negative impacts on firms’ business activities. The innovation compensation effect did not significantly outweigh the cost effect, and thus CACER failed to promote economic growth in China. MBER allows enterprises to consider the cost-benefit balance according to their own business conditions and seek more cost-effective pollution control and emission reduction methods. This quest for cost-effectiveness has prompted firms to seek more innovative, energy-efficient, and environmentally friendly technologies and processes. Thus, MBER has contributed to China’s economic growth.

4.2. Regional Heterogeneity

China is a vast country with marked differences in its natural environment. Factors such as resource allocation and climatic conditions may have an impact on economic activities and social development in each region. Moreover, there may be differences in the market characteristics of different regions, including market size, and degree of competition. Referring to the grouping in the China Statistical Yearbook, this paper categorizes China’s 30 provinces into two groups: the eastern and central regions, and the western region. The eastern and central region includes 19 provinces, and the western region includes 11 provinces.

In general, eastern and central China is more developed and western China has a lower level of development. The western China, in particular, is developing slowly despite its rich natural resources. The impact of environmental regulation on economic growth may be subject to many conditions and exhibit non-linear or heterogeneous characteristics. To test this conjecture, this paper starts with a regional heterogeneity analysis.

Table 3. The regress results of regional heterogeneity.

	Eastern and Central	Western	Eastern and Central	Western
MBER	7.434 **	0.146
	(1.98)	(0.04)
CACER			1.012	−0.931 *
			(0.97)	(−1.67)
IS	8.756 **	5.917	9.972 **	5.253
	(2.11)	(1.18)	(2.35)	(1.17)
OPEN	0.560 ***	0.165	0.554 ***	0.100
	(5.04)	(0.70)	(4.92)	(0.42)
ST	−233.893 **	106.287	−264.156 **	71.922
	(−2.21)	(0.75)	(−2.50)	(0.51)
PCI	−2.100 **	2.075 ***	−2.007 **	2.273 ***
	(−2.26)	(2.63)	(−2.14)	(2.89)
Constant	5.578 ***	7.294 ***	5.549 ***	8.119 ***
	(2.77)	(2.75)	(2.65)	(3.78)
Observations	247	143	247	143
Fixed effects	YES	YES	YES	YES
R2	0.814	0.910	0.750	0.905

(t value in parentheses, *** p < 0.01, ** p < 0.05, * p < 0.1).

reports the regress results of regional heterogeneity.

For MBER (market-based environmental regulation): In the eastern and central regions of China, the regression coefficient of MBER is significantly positive, while in the western region, the regression coefficient of MBER is not significant. The impact of MBER on economic growth is not the same in different regions: In the eastern and central regions, MBER has facilitated economic growth, while in the western region, MBER can not promote economic growth. The possible explanation is that the innovation effect of market-based environmental regulation needs to be supported by a well-developed technical market and science and technology industry chain. The eastern and central regions of China are relatively economically developed, with a well-developed industrial structure and technology industry chain. In these regions, strict MBER can encourage enterprises to increase technological innovation, improve resource utilization efficiency and productivity, and thus promote sustainable economic growth. In addition, the market demand in the eastern and central regions is strong, and enterprises are more willing to invest in innovation in order to meet the market demand and improve market competitiveness. In western China, the inhibitory effect of MBER on economic growth is mainly due to the region’s relatively lagging economic development, a single industrial structure, and an imperfect science and technology industry chain. In addition, the relatively weaker market demand and less market pressure on firms in the western region may have reduced their incentives to innovate. Therefore MBER can not promote economic growth in the western region.

For CACER (command-and-control environmental regulation): In the eastern and central regions of China, the regression coefficient of CACER is not significant, while in the western region, the regression coefficient of CACER is significantly negative. The impact of CACER on economic growth is not the same in different regions. In the eastern and central regions, CACER has no impact on economic growth, while in the western region environmental regulations hinder economic growth. The probable reason is that the underdeveloped economy of the western region and the lack of a well-developed science and technology industry chain and innovative market demand further exacerbate the cost effect and weaken the innovation effect, thus leading to CACER, which inhibits economic growth.

4.3. Nonlinear Regression

Numerous studies have also discovered that there is a nonlinear relationship between environmental regulation and economic growth, with the impact of environmental regulation on growth changing with its level of intensity [27,45]. A U-shaped association is still observed in some literature, despite the majority of the evidence supporting an inverted U-curve relationship [46]. Therefore further validation is needed.

According to the previous analysis, the impact of environmental regulation on economic growth depends on the relative size of the innovation effect and cost effects, thus may exhibit a U-shaped or inverted U-shaped curve. In order to test Hypothesis 2 and to test whether the impact of environmental regulation on economic growth varies with changes in environmental regulation, the paper further adds a quadratic term for environmental regulation to Equations (1) and (2) and get the following model:

Growth_it = α + β₁MBER_it + β₂MBER²_it + β₃CON_it + d_Province + d_year + ε_it

(3)

Growth_it = α + β₁CACER_it + β₂CACER²_it + β₃CON_it + d_Province + d_year + ε_it

(4)

The linear regression results of the baseline model show that, overall, MBER significantly increases China’s economic growth So does MBER regulation keep having a positive impact on China’s economic growth, or does this positive impact disappear once environmental regulation reaches a certain level? Similarly, The linear regression results of the baseline model show that, overall, the CACER can not promote China’s economic growth. So, does CACER also exhibit a U-shaped or inverted U-shaped curve?

Table 4. Nonlinear regression results.

Variables	(1)	(2)	(3)	(4)
MBER	34.176 ***	35.498 ***
	(6.18)	(6.39)
MBER2	−91.831 ***	−90.573 ***
	(−6.06)	(−6.02)
CACER			1.837	2.164 **
			(1.18)	(2.01)
CACER2			−1.001	−1.031
			(−1.45)	(−1.57)
IS		9.536 ***		8.598 *
		(2.95)		(1.78)
OPEN		0.272 ***		0.336 *
		(2.82)		(1.74)
ST		10.495		−61.398
		(0.13)		(−0.41)
PCI		−0.382		−0.694
		(−0.60)		(−0.47)
Constant	8.811 ***	3.637 **	10.406 ***	6.016 **
	(18.73)	(2.18)	(15.07)	(2.57)
Observations	390	390	390	390
Fixed effects	YES	YES	YES	YES
R2	0.757	0.768	0.761	0.774

(t value in parentheses, *** p < 0.01, ** p < 0.05, * p < 0.1).

reports the regression results. The first column reports the results of MBER (market-based environmental regulation) without the control variables and the second column reports the results of MBER with the control variables. The MBER is significantly positive and MBER² is significantly negative. This shows that the relationship between MBER and economic growth is an inverted U-shaped curve: MBER promotes economic growth at low levels of intensity, but beyond a certain point, it becomes a barrier to economic growth. MBER causes two effects: a cost effect and an innovation compensation effect. Thus, if the cost effect is greater than the innovation compensation effect, then the impact of environmental regulation on economic growth should be shown to be inhibitory. If the innovation compensation effect is greater than the cost effect, then MBER should promote economic growth. MBER has a greater innovation effect than a cost effect at lower intensities and a greater cost effect than an innovation effect at higher intensities and therefore this results in an inverted U-shaped curve. Considering the low intensity of MBER in China as a whole, It is consistent with the positive average effect obtained from the linear model. The following analysis in this paper is still based on the linear model to discuss the average impact of MBER on economic growth.

For MBER, Hypothesis 2 is proven. The scatter plot and fitting curve of MBER and economic growth are presented (Figure 2). The scatter distribution of each variable reveals the following: The majority of the sample points for MBER are clustered to the left of the symmetry axis. This pattern indicates that, despite the regression results suggesting an inverted U-shaped relationship, market-based environmental regulation policies continue to exert a positive impact on overall economic growth. Consequently, the non-linear analysis supports the baseline regression findings, affirming the robustness of the model.

The third column presents the results of CACER (command-and-control environmental regulation) without control variables, whereas the fourth column presents the results of CACER with control variables. The regression coefficient of CACER² is not significant, which demonstrates that there is no U-shaped or inverted U-shaped curve relationship between economic growth and CACER. For CACER, Hypothesis 2 is not proven. Thus the impact of CACER on economic growth still exhibits only a linear relationship: according to the baseline results in Table 2, CACER has no effect on economic growth and does not promote economic growth.

5. Robustness Check

5.1. Partial Sample Regression

In 2018, pollution charges were replaced by environmental protection taxes. The pollutants levied by the pollution charge and the environmental protection tax are the same. The pollution charge and environmental protection tax are in the same lineage, so these two data are used simultaneously. But to ensure the robustness of the results, this paper uses only pollution charges (2009–2017) in the regression.

In addition, due to COVID-19, economic growth in 2020 and 2021 is characterized differently from the other years. As a result of the restrictive measures implemented to contain the spread of the epidemic, these measures affected various industries and sectors to varying degrees, thereby affecting overall economic growth. We therefore exclude these two years and re-run the regression using data from 2009 to 2019.

Table 5. Partial sample regress.

Variables	MBER		CACER
	2009–2017	2009–2019	2009–2019
MBER	6.813 *	8.228 ***
	(1.70)	(2.60)
CACER			0.666
			(1.17)
IS	6.074	9.292 **	8.498 **
	(1.41)	(2.58)	(2.34)
OPEN	0.571 ***	0.439 ***	0.440 ***
	(4.14)	(4.07)	(4.04)
ST	−49.009	−10.766	−25.943
	(−0.39)	(−0.11)	(−0.27)
PCI	−0.257	−1.088	−1.149
	(−0.26)	(−1.53)	(−1.60)
Constants	6.145 ***	5.140 ***	6.277 ***
	(2.93)	(2.91)	(3.66)
Observations	270	330	330
Fixed effects	YES	YES	YES
R2	0.719	0.744	0.739

(t value in parentheses, *** p < 0.01, ** p < 0.05, * p < 0.1).

reports the regression results. The results remain the same as in the baseline regression.

5.2. Endogeneity

(i)

Mitigating the problem of “omission bias”

Considering the possible important roles of education inputs, level of economic development, and industrial agglomeration in economic growth rate, based on Equations (1) and (2), this paper further controls the factors of education inputs and level of economic development in economic growth rate that may affect economic growth in the regression so as to alleviate the possible “omission bias” problem. Based on the consideration of data availability, this paper selects “the ratio of fiscal expenditure on education to GDP” to measure educational inputs (EI), and “Per capita GDP (log)” to measure the level of economic development (ED). In general, the higher the level of economic development (GDP per capita), the lower the rate of economic growth. In addition, referring to the practice of Wang et al. [47], this paper measures the level of industrial agglomeration (IA) and its calculation formula is IA_i = (q_i/q)/(Q_i/Q). q_i denotes the number of employees in the manufacturing industry in province i, and q denotes the number of employees in the manufacturing industry in the country. Q_i denotes the total number of employees in the province i, and Q denotes the total number of employees in the country.

Table 6. Mitigating the problem of omission bias.

Variables	(1)	(2)	(3)	(4)	(5)	(6)
MBER	5.856 **	4.839 *	5.883 *
	(2.06)	(1.68)	(1.94)
CACER				0.811	0.619	0.644
				(1.42)	(1.08)	(1.12)
IS	9.147 ***	12.822 ***	6.365	9.189 ***	13.099 ***	9.731
	(2.61)	(3.19)	(0.90)	(2.59)	(3.24)	(1.42)
OPEN	0.341 ***	0.387 ***	0.373 ***	0.343 ***	0.392 ***	0.386 ***
	(3.40)	(3.76)	(3.60)	(3.41)	(3.80)	(3.71)
ST	−34.429	−16.863	−34.997	−56.151	−33.142	−44.308
	(−0.40)	(−0.19)	(−0.40)	(−0.65)	(−0.38)	(−0.50)
PCI	−0.868	−0.500	−0.235	−0.966	−0.547	−0.407
	(−1.31)	(−0.72)	(−0.32)	(−1.45)	(−0.78)	(−0.55)
EI	26.006	−0.437	−1.811	36.174	5.613	−2.166 *
	(1.03)	(−0.02)	(−1.51)	(1.43)	(0.19)	(−1.83)
ED		−2.139 *	−0.730		−2.304 **	5.755
		(−1.84)	(−0.03)		(−1.99)	(0.20)
IA			2.598			1.361
			(1.11)			(0.61)
Constant	4.782 **	25.366 **	22.492 *	4.917 **	27.098 **	25.971 **
	(2.31)	(2.24)	(1.93)	(2.34)	(2.39)	(2.26)
Obs	390	390	390	390	390	390
Fixed effects	YES	YES	YES	YES	YES	YES
R2	0.745	0.746	0.747	0.774	0.777	0.777

(t value in parentheses, *** p < 0.01, ** p < 0.05, * p < 0.1).

reports the regression results of mitigating the problem of omission bias. The findings of the regression demonstrate that the MBER’s coefficients are still significantly positive, and the estimation results did not change significantly compared with the baseline results, which indicates that the promotional effects of MBER on the economy are not affected by the bias of the omitted variables. The coefficients of the CACER are still not significant, and the estimation results did not change compared with the baseline results.

(ii)

Two-stage least squares estimation

To further alleviate the endogeneity concerns, we performed a two-stage least squares (2SLS) estimation. Referring to the practice of Yin et al. [40], this paper uses lagged one-period values of MBER and CACER as instrumental variables, considering environmental regulation as an endogenous variable. The efficiency of the instrumental variables is demonstrated by their passing both the weak instrumental variable test and the unidentifiable test. The endogeneity problem of “reverse causality” is somewhat mitigated by the fact that environmental regulations in the lagged period are predetermined and unaffected by current shocks, making them uncorrelated with the error term and devoid of reverse causality with them, such as economic growth in the current period.

Table 7. The regression results of two-stage least squares estimation.

Variables	(1)	(2)
MBER	22.843 ***
	(4.23)
CACER		3.040
		(1.57)
IS	33.219 ***	30.819 ***
	(13.93)	(11.06)
OPEN	0.383 ***	0.470 ***
	(3.00)	(3.63)
ST	−169.341	−255.435 **
	(−1.56)	(−2.27)
PCI	−3.294 ***	−4.387 ***
	(−4.36)	(−4.54)
Constant	−1.006	2.025
	(−0.59)	(1.29)
Observations	360	360
Fixed effects	YES	YES
Kleibergen Paap rk LM	30.724 ***	11.069 ***
Kleibergen Paap rk Wald F	158.939 ***	27.741 ***

(t value in parentheses, *** p < 0.01, ** p < 0.05. Kleibergen Paap rk LM is an unidentifiable test statistic, while Kleibergen Paap rk Wald F is a weak instrumental variable test statistic.).

reports the regression results of two-stage least squares estimation. The coefficient estimate on MBER remains significantly positive, suggesting that after accounting for endogeneity issues, in general, MBER in China has contributed to China’s economic growth. The coefficient estimate on CACER remains not significant, suggesting that after accounting for endogeneity issues, CACER in China can not promote China’s economic growth.

6. Further Analysis of Market-Based Environmental Regulation

The results of the baseline regression indicate that MBER promotes economic growth while CACER does not. Nonlinear regression shows that the relationship between MBER and economic growth is an inverted U-shaped curve. However, CACER does not exhibit nonlinear characteristics. MBER is superior to CACER in terms of economic growth effects. The purpose of this paper is to investigate how to achieve a win–win situation between environmental protection and economic growth. It is therefore necessary to further clarify the constraints on MBER promoting economic growth. Consider that only MBER promotes economic growth, while CACER does not. Therefore, the analyses that follow in this paper are based on MBER.

6.1. Impact of Level of Economic Development

The environmental Kuznets curve shows that the relationship between the level of economic development and pollution changes as the level of economic development changes [48]. Then, would the impact of environmental regulation on economic growth also be constrained by the level of economic development? If the cost effect of environmental regulation does exist, it becomes more apparent at lower levels of economic development. This is because in regions with low levels of economic development, firms tend to be more underfunded. Financial markets may be underdeveloped, making it difficult for firms to obtain financial support. Even a low intensity of environmental regulation crowds out funds for routine economic development, thus seriously hampering the normal production and operation of firms.

To test Hypothesis 3, based on the size of the annual average per capita GDP from 2009 to 2021, China’s thirty provinces were split into two categories in this paper: those with high economic development and those with low development. This further verifies whether the impact of MBER on economic growth is constrained by the level of economic development. Economically high-development provinces include Beijing, Shanghai, Tianjin, Jiangsu, Zhejiang, Fujian, Guangdong, Inner Mongolia, Shandong, Liaoning, Chongqing, Hubei, Shaanxi, Jilin, and Hunan. Economically low-development provinces include Ningxia, Hainan, Xinjiang, Hebei, Henan, Anhui, Qinghai, Sichuan, Jiangxi, Shanxi, Heilongjiang, Guangxi, Yunnan, Guizhou, and Gansu. Based on Equation (1), this paper conducts regressions for the two sample groups separately.

Table 8. Impact of the level of economic development.

Variables	Provinces with High Levels of Development	Provinces with Low Levels of Development
MBER	25.269 ***	−6.368 **
	(4.31)	(−2.07)
IS	2.425	3.067
	(0.37)	(0.86)
OPEN	0.447 **	0.118
	(2.58)	(0.88)
ST	−168.736	175.990
	(−1.42)	(1.53)
PCI	−1.984	0.184
	(−1.29)	(0.27)
Constant	7.805 **	9.782 ***
	(2.60)	(4.96)
Observations	195	195
Fixed effects	YES	YES
R2	0.763	0.860

(t value in parentheses, *** p < 0.01, ** p < 0.05).

provides the regression results of the subgroup regression based on Equation (1). The first column shows the results of the regression for provinces with high levels of development, where the regression coefficient for MBER is positive significantly. The second column shows the regression results for provinces with low levels of development, where the regression coefficient of MBER is significantly negative. According to the regression analysis, MBER hinders economic growth when the economic development level is low, whereas MBER acts as a stimulant for growth when the economic development level is high.

In order to further explore the inflection point accurately, i.e., what level of economic development (GDP per capita) is required for market-based environmental regulation to promote economic growth, referring to Hansen [49], this paper constructs the following threshold model:

Growth_it = β₀ + β₁MBER_it × I(PGDP ≤ TV) + β₂MBER_it × I(PGDP > TV) + β₃Control + ε_it

(5)

PGDP represents the threshold variable: per capita GDP. I(.) is an indicative function. when PGDP ≤ TV, I = 1; otherwise I = 0. TV is the threshold value. Control variables are consistent with the previous section.

According to Hansen [49], the threshold effect of the model is first tested. Using real GDP per capita as the threshold variable, Equation (5) is estimated under the original hypotheses of single thresholds in turn, and the F-statistic and the p-value derived using the bootstrap method.

Table 9. Results of the market-based environmental regulation threshold effect test.

Threshold Type	F Stat Value	p Value (H0: No Threshold)	BS	Threshold Value
Single	53.54	0.00	500	85,422

reports the test results.

The test results show that the threshold test was passed at the 1% level of significance, and thus there is a single threshold effect. The results of the threshold model are as follows:

Table 10. Regression results of the threshold model.

Variables	Economic Growth Rate
MBER(PGPD ≤ TV)	−3.843
	(−1.28)
MBER(PGDP > TV)	38.741 ***
	(7.23)
Control variables	YES
Observations	390
R-squared	0.803
Number of provinces	30

(t value in parentheses, *** p < 0.01).

reports the regression results of the threshold model. The regression coefficient of MBER(PGPD ≤ TV) is not significant, while the regression coefficient of MBER(PGDP > TV) is significantly positive. MBER has no effect on economic growth when the level of economic development is below the threshold. When the level of economic development is above the threshold, MBER promotes economic growth. This result proves Hypothesis 3.

Because innovation needs to be supported by sufficient funds, well-established infrastructure, and mature markets, only economically developed provinces or regions have such conditions. If the economy is underdeveloped, the innovation effect of MBER might not be fully realized, potentially impacting the relationship between MBER and economic growth. The threshold model can accurately estimate the numerical magnitude of the turning point. The estimated value of the inflection point can be compared with the current state of China’s economic development (GDP per capita), which has practical significance: different provinces/regions should choose the appropriate intensity of MBER according to their level of economic development.

According to the threshold value, market-based environmental regulation is divided into two intervals: GDP per capita less than 85,422 CNY (Approximately 12,000 USD) and more than 85,422. The regression coefficient of market-based environmental regulation only when GDP per capita is more than 85,422 CNY is significantly positive. It shows that MBER promotes economic growth only when the GDP per capita is more than 85,422 CNY.

6.2. Impact of Financial Development

For firms, environmental regulation imposes an additional institutional cost, and how to defuse that cost has a bearing on the stable functioning of the real economy as well as on sustainable development. Typically, enterprises can reduce emissions by reducing output or tackling pollution. However, pollution control requires environmental investment or technological innovation, which are generally characterized by high risk, low up-front returns, and long lead times, making environmental investment and pollution control by firms likely to suffer from external financing constraints and only respond to environmental regulations by reducing output. Financial development can ease financing constraints for enterprises and lower the costs associated with environmental investment and financing, thereby enhancing the effectiveness of environmental regulation policies and stimulating economic growth. Additionally, from the perspective of scientific and technological research and development through to the commercialization of innovations, and then to the development of the market and the generation of benefits, each stage requires a large amount of capital investment. If enterprises are subject to financial constraints, it will limit their technological innovation. Finance, as a hub of resource allocation, can provide financial support for scientific and technological innovation, and can also provide financial tools to avoid innovation risks, which is an important guarantee for scientific and technological innovation.

To test Hypothesis 4, this paper introduces financial development variables to Equation (1) and builds the following model:

Growth_it = α + β₁MBER_it + β₂FD_it + β₃MBER_it × FD_it + β₄CON_it + d_province + d_year + ε_it

(6)

where FD_it represents the financial development in year t.

Since the concept of financial deepening was introduced by Shaw [50] and McKinnon [51], a large body of literature has used “total financial assets as a percentage of regional GDP” to measure the level of financial deepening or monetization in a region [52,53]. Following the literature, this paper uses the financial deepening rate indicator (FD), which is the ratio of loan balances to GDP, to measure the financial development of provinces.

Table 11. Impact of financial development.

Variables	(1)	(2)
MBER	−15.734 **	−9.522
	(−2.00)	(−1.19)
FD	−0.874 ***	−0.679 **
	(−3.05)	(−2.31)
MBER×FD	14.193 ***	10.915 **
	(2.79)	(2.10)
IS		7.869 **
		(2.32)
OPEN		0.295 ***
		(2.88)
ST		6.545
		(0.08)
PCI		−0.575
		(−0.86)
Constant	11.897 ***	7.235 ***
	(17.97)	(4.07)
Observations	390	390
Fixed effects	YES	YES
R2	0.768	0.778

(t value in parentheses, *** p < 0.01, ** p < 0.05).

reports the regression result. In both the first and second columns, the coefficients for MBER×FD are significantly positive. This indicates that the higher the level of financial development in a province, the more pronounced the impact of MBER on economic growth. Therefore, financial development can synergistically enhance the effectiveness of MBER in driving economic growth. This result proves Hypothesis 4.

7. Conclusions

While China’s economic development has achieved remarkable success, it has also paid a heavy price in terms of the environment, which has led to a series of social problems related to environmental pollution. Protecting the environment is an inevitable policy choice for the government. Environmental issues are the hotspots and focal points of China’s economic development, and effectively balancing environmental protection with economic growth is crucial for sustainable development. This paper investigated the economic impacts of environmental regulation, took into account China’s current situation, and explored whether environmental regulation is likely to promote economic growth.

This paper explored the relationship between economic growth and environmental protection using panel data for 30 provinces in China from 2009 to 2021. First, the regression results showed that MBER promotes economic growth. However, CACER in China can not promote China’s economic growth. Second, further research has shown that environmental regulation’s role in promoting the economy is constrained by the intensity of environmental regulation. MBER is only able to promote economic growth when the intensity of MBER is low. If the intensity of MBER is too high, MBER, on the contrary, can not promote economic growth. Third, The impact of MBER on economic growth also was found to be constrained by the level of economic development: the MBER does not promote economic growth when the economy is less developed. It is only when the economy has reached a high level that market-based environmental regulation will contribute to economic growth. Finally, this paper found that financial development and MBER can synergize to promote economic growth.

Several recommendations are provided in this study based on these findings. Firstly, environmental protection and economic growth can be a win–win situation, and in general, China should still strengthen environmental protection. However, the appropriate environmental regulatory tool should be selected. As far as the economic growth effect is concerned, the choice of market-based environmental regulatory instruments should be implemented, while the implementation of command-and-control environmental regulations should be reduced. Nonetheless, the degree of environmental protection ought to be assigned based on the conditions in various areas. In China’s eastern and central areas as well as in the economically developed provinces, market-based environmental regulation can continue to be strengthened. In the western and economically underdeveloped provinces, overly stringent environmental regulations should be avoided. Finally, Chinese provinces should strengthen their financial support in order to better utilize the role of market-based environmental regulation in promoting economic growth.

Finally, what need to be mentioned are the limitations of this paper. There are a variety of ways to measure market-based and command-and-control environmental regulation and this study simplified the impacts of different types of environmental regulations on economic growth, which may not fully reflect the complexities of real-world economic systems. In addition, there are many socio-economic differences between different provinces and regions in China, and due to data constraints, this paper only considered these differences from a limited number of perspectives and also focused mainly on short- and medium-term impacts, ignoring long-term impacts. Future research should incorporate more detailed factors in order to gain more comprehensive insights.