Financial Decentralization and Environmental Pollution Research—An Empirical Test Based on Data from 279 Cities in China

Abstract

By establishing a principal–agent model, this study introduced the local government’s behavior and deduced the local government in the face of incentive policies, which means that it has more power to develop the economy at the cost of environmental pollution. Furthermore, from the perspective of fiscal expenditure structure and local government competition, the influence path of fiscal decentralization on environmental pollution is examined. Thus, this methodology unifies three seemingly different but strongly interrelated hypotheses into one single composite model. Firstly, the bidirectional fixed effect model tests the relationship between fiscal decentralization and environmental pollution empirically and shows that the improvement of fiscal decentralization will aggravate environmental pollution. Additionally, combined with the inverted U-shaped relationship between economic development and environmental pollution, this paper argues that fiscal decentralization has different impacts on environmental pollution at different levels of economic development. Secondly, the results of the mechanism test prove that fiscal decentralization will increase regional competition, increase the scale of foreign capital utilization, and reduce the proportion of local government expenditure for public welfare, thus increasing environmental pollution. Thirdly, heterogeneous regression results demonstrate that fiscal decentralization will increase the environmental pollution of high pollution areas and non-provincial capital cities but has no significant impact on low pollution areas and provincial capital cities. Based on the above conclusions, this paper develops countermeasures to control environmental pollution from optimizing fiscal decentralization structure.

1. Introduction

As China’s economic growth has entered the new standard, environmental pollution control has become one of China’s three existing critical battles. After more than 40 years of reform and opening-up, China’s economy reached the second in the world, and its growth is increasingly contributing to the world economy. However, rapid urbanization and industrialization in China are accompanied by severe environmental pollution problems. For example, in the 2020 World Environmental Quality ranking, China only ranks 103rd, and smog and environmental emergency incidents are constantly emerging, bringing significant challenges to China’s economy’s sustainable development. At the present stage, the Party and the government pay great attention to environmental pollution, strengthen the construction of ecological civilization, and adhere to the development concept of “innovation, coordination, green, open, sharing”, which is an essential task of China’s environmental protection at the present stage. However, due to the long realization of extensive economic development mode in China, the causes of environmental pollution are more complex. Therefore, clearing the influencing factors of environmental pollution in China and taking targeted countermeasures is the primary task of pollution prevention and control.

Since 1994, the tax-sharing reform has been carried out between the central government and local governments. Since then, the main factor in China’s sustained economic growth has been the local government competition. Furthermore, achieving all this is financial decentralization. It is worth noting that although the local government competition represented by the “GDP” championship has promoted the rapid growth of China’s economy, it has also brought severe environmental pollution problems. For example, environmental pollution, soil erosion, environmental damage, and other natural problems cause ecological pollution, but the economy cannot be preserved. “Three industrial wastes” have brought air pollution, water pollution, haze pollution, and other problems that have also seriously affected the lives of residents, and sudden environmental emergencies have been staged in various regions. Environmental pollution poses higher challenges to improving the quality of China’s economic development and quality of life. We must value our environmental issues. Otherwise, the consequences will become severe, and social stability will be compromised. Ecological issues will become social problems or even political issues. Therefore, China’s arduous tasks and challenges will still be environmental problems in the social transition period. Thus, whether we can protect the clear waters and green mountains is our challenge.

Under today’s performance evaluation mechanism, GDP is the performance of local officials. With the strengthening of China’s attention to environmental pollution, environmental concerns have been a wide concern by all walks of life in recent years. The environmental pollution problem is relatively extreme, which seriously impacts the quality of China’s economic development. Under the decentralization system, the local financial pressure has intensified the difficulty of environmental governance. Consequently, under the background of the current financial decentralization system, combined with the actual needs of China’s economic development, it is of great significance to study the influence of financial decentralization on environmental pollution for the sustainable development of China’s economy.

This paper has a rich theoretical basis. Firstly, by exploring how financial decentralization affects the environment, and the impact of the economic construction of local governments, and then conducting a substantive test. This article can fully introduce the association between fiscal decentralization and ecological pollution and explore their relationship. Thus, the countermeasures are developed. In addition, different from previous studies, introducing the role of local government, studying the motivation of local government under the existing government management performance evaluation, and clarifying its impact on environmental pollution can also expand the scope of research in the field of government governance.

Environmental problems are the focus of modern economics research. The biggest innovation of this paper lies in the introduction of local government behavior. By establishing a principal–agent model, it deduces the choice made by local governments in the face of central incentive policies; that is, they have more motivation to develop economy at the cost of environmental pollution. In addition, in the empirical study, this paper uses city-level data to explore the impact of fiscal decentralization on environmental pollution, and studies the impact path of fiscal decentralization on environmental pollution from the perspective of fiscal expenditure structure and local government competition, so as to achieve the combination of theoretical research and empirical testing.

At the same time, it should also be recognized that due to data limitations, this paper can only measure environmental pollution from the three industrial wastes. Due to the lack of data, some urban samples are screened out, and some control variables cannot be quantified. In terms of mechanism, we can only start from the perspective of financial expenditure structure and government competition. In fact, there are more local government actions that cause environmental pollution. In addition, in the choice of control variables, the industrial structure of heavy industry, climate variables, the scale of emissions trading, and environmental protection facilities into severe degree and environmental monitoring are also important factors, but due to the limitation of data availability, this article does not add the above variables to the regression model. This is a limitation of this paper, and is also the future direction for further study.

2. Literature Review

Financial decentralization is an agreement between the central and local governments to allocate monetary funds [1]. Strict conditions are required to make fiscal decentralization to the government form positive incentives and promote the implementation of appropriate policy. Otherwise, fiscal decentralization will bring unexpected results, including macroeconomic fluctuations and low economic growth, including the lack of public services, unequal allocation of resources between regions, and even government corruption [2,3]. Lin [4] believes that under China’s unitary government system, the authoritarian central government can bring strong incentives to local governments and improve local governments’ management efficiency by cleverly setting up the “decentralization” mechanism. However, due to the incredible complexity of political organizations in incentives, it is not easy to obtain accurate measures, so the government chose GDP growth as an indicator to examine the performance of government officials [5].

As for the expense of financial decentralization, scholars focus mainly on the unbalanced economic development, the supply of public services, corruption, and the gap between urban and rural development [6]. For example, Hu et al. [7] assume that the cities in the Yangtze River Delta and Pearl River Delta have a high level of financial decentralization, thus obtaining more foreign capital and resource allocation and acquiring more performance in the economy and trade, accordingly, aggravates the regional imbalance of China’s economic development. Cheng et al. [8] pointed out that the level of financial decentralization in developed areas in China is relatively higher, which can drive the vitality of regional economic development but will bring pressure on the coordinated development of the whole regional economy in the demonstration. In addition, Liu et al. [5] believe that the government tends to reduce the proportion of agriculture and public service expenditure under the dual constraints of political competition and a decentralization system.

In different countries, local governments are under different constraints due to the competition of environmental rules, resulting in the impact of environmental quality being different. Policies on ecological governance are also different. An et al. [9] pointed out that the promotion of local officials is essentially a zero-sum game. This is due to the domestic political ecology. Huang [10] believes that because it is tough to collect data, the role of environmental regulations cannot be objectively evaluated, and it is also challenging to examine with experience. Yang et al. [11] argue that local finance is responsible for the funding for local environmental protection departments at all levels and the formulation and dismissal of local environmental protection. Local governments advocating for environmental protection will slow down economic development [12]. Local governments are increasing environmental regulation, which means they will implement environmental protection and management policies. The policies created by the central government will improve the local ecology, while the negative effect will impact investment and economic growth.

Liu et al. [13] consider that because local governments are keen on economic growth in the competition, they will maximize investment attraction. This intergovernmental pursuit of foreign investment infringes on environmental quality and weakens environmental quality standards. Wu [14] pointed out that the higher the decentralization of local finance, the more restrictive policies local governments have on environmental preservation.

In the existing research conclusions of comprehensive scholars can be found that the environmental pollution problem in economic development has been a wide concern. The most widely used curve is the environmental Kuznets curve, which comprehensively describes the environmental pollution quality problem under different economic development states. Nevertheless, due to the solid negative externalities of environmental pollution, the motivation of economic individuals to actively control pollution is insufficient, which determines that the government will play an essential role in environmental governance in China. In China’s market economy construction, governments at all levels need to provide more robust governance measures to deal with the environmental pollution problem caused by the market economy failure. Financial decentralization is an important measure to coordinate the financial power of the central and local governments and promote local economic development. Some existing academic studies have proved that fiscal decentralization has adverse effects on environmental pollution. However, the empirical studies of scholars mainly conduct a simple empirical analysis to test the linear relationship between the two, but they do not reveal the mechanism of action. Therefore, from the perspective of local competition and financial expenditure, this paper empirically examines the role mechanism of fiscal decentralization and environmental pollution and provides relevant policy suggestions for optimizing China’s fiscal decentralization policy and playing the role of fiscal decentralization in the economy and market.

3. Theoretical Analysis

3.1. Local Government Behavior under the Decentralized System

3.1.1. Differences in the Expected Targets between Central and Local Governments

The theory of fiscal decentralization originates from the West. The traditional fiscal decentralization theory represented by Tiebout’s “voting with feet” theory [15] and Musgrave’s decentralization theorem [16] holds that, compared with the central government, local governments understand the preferences and utility of local residents better, and the provision of public goods by local governments will be more effective. Therefore, fiscal decentralization will improve the quality of the environment. However, the traditional theory of fiscal decentralization has a series of assumptions, one of which is that the government is selfless. However, the new generation of financial theories represented by Qian et al. [17] and Wildasin [18] revised this assumption, introduced the theory of mechanism design and mechanism compatibility into the traditional theory of fiscal decentralization, denied the premise of government selflessness, and believed that government officials would pay attention to their own interests. Once there was no restriction, officials would try every means to seek rent in political decision-making. Therefore, compared with the traditional theory of fiscal decentralization, The new generation of fiscal decentralization theory is more in line with the current Chinese fiscal decentralization system, and can better study the relationship among fiscal decentralization, local government competition and environmental pollution.

China’s local governments are facing both economic and political incentives. Under this circumstance, the local government’s resource investment will be related to the local government’s interests under the incentives of both sides. Thus, investing resources in projects can bring more significant benefits to themselves in the short term. Overall, represented by infrastructure investment, industrial development projects in the short term can bring more incredible economic benefits, provide more convenience for officials’ promotion, consequently, control the local financial resources of local governments, will concentrate on financial resources for more attractive investment projects, and ignore the essential public services.

Through the current political incentive mechanism of financial decentralization, governments at all levels regard the growth of regional GDP as a vital assessment standard for local officials, and the “GDP only theory” exists widely among local governments. After financial decentralization, the horizontal expansion of local government in the regional economic competition is strengthened, and the competition degree of local government is intensified so that the expenditure scale of economic construction is too large. At the same time, the environmental protection problem is ignored. Under the existing political incentive mechanism of financial decentralization, local governments can obtain a certain amount of tax revenue from the local economic development to a certain extent through the tax distribution degree. Under the incentive of tax revenue, local governments will convey stronger local competition to increase their income. At the level of fiscal expenditure, the tax sharing system brings about the difference between the central and local administrative power and the taxing power, resulting in the fiscal revenue being lower than the expenditure it spends. This gap needs to be filled, and the local governments have to vigorously develop the local economy and expand the local tax sources. Some enterprises with high pollution and high energy consumption are often large local taxpayers, which brings about more significant environmental pollution concerns.

From what has been discussed above, this article will argue that the main factors of the government’s target utility function include local economic level, fiscal revenue, and official consumption. Figure 1 shows a schematic diagram of the government in the existing financial decentralization system. Through the literature research [19,20], the author uses the following functions to express the government utility:

$$U_i=\alpha ln{x}_i+\beta ln{y}_i+\lambda ln{z}_i+\epsilon$$

(1)

In the local government utility function (1), i means the i th local government, x represents the development level of the local economy, y indicates the concern of the next level of government on the economic development, and generally speaking, it is significantly positive’s said that the fiscal revenue of local governments, for local governments, the larger the fiscal revenue, the more conducive to the function of the government, so it should be significantly positive. It represents the personal interests of local officials, similar to the phenomenon of principal and agency in economic theory. It is worth noting that, with the degree of emphasis. The more attention the central government attaches to the GDP of each region, the higher its value is. Considering the current domestic mechanism, the author sets the function as:

$$U_i={{\displaystyle \int }}_1^t{U}_i\left(x_i,y_i,z_i\right)e^{-pt}d\left(t\right)$$

(2)

In the target function during the term (2), t indicates the time of the local government official, p is the discount rate during the term, which is constant. In this function, the improvement of local economic level, the expansion of fiscal revenue scale and the increase of official income can enhance the overall utility of local officials, and $u_x^{\prime }>0, u_y^{\prime }>0, u_z^{\prime }>0$. Due to the limited term of government officials, local officials tend to develop local economies in order to maximize their utility in the short term. As financial resources are centrally managed, local governments have a greater incentive to conduct economic interference through fiscal revenue and expenditure, and the improvement of the level of financial decentralization will increase such interference. Due to the different political and economic aspects, lead to place in the process of policy implementation will be not fully achieve the central set goals, lead to policy implementation of distorted economic behavior, when financial means are mainly used for economic construction, represented by environmental protection of public welfare undertakings will be shelved, inevitably bring pollution caused by ignoring environmental protection.

3.1.2. Environmental Governance by Unpowered Local Governments

Due to the administrative system in which China is responsible to the upper level but not to the lower level, the difference between the utility function of the central and local governments is brought about. This implementation difference between the central and local governments leads to significant tension when governing the environment. Under the system of financial decentralization, the central government often formulates a series of environmental protection rules and regulations in the process of economic development to reduce the discharge of pollutants. As mentioned above, local governments pay more attention to short-term interests. Therefore, local governments have a stronger motivation to carry out the “GDP-only theory” and assume more leadership roles in economic construction in regional governance. In recent years, governments have developed the local economy, increased the scale of foreign investment and capital introduction, relaxed the environmental supervision, and reduced the regional pollution entry threshold. Under this background, some high pollution, high energy consumption enterprises have been established, bringing excess capacity and increasing the scale of pollution emissions. Because of the lack of funds and a wide range of pollution, environmental problems are in the process of no power. Under the background of tax sharing reform, the financial resources of local governments are tightly tied up by the central government, and the fiscal revenue of local governments is reduced, but the relevant local management affairs have not been reduced accordingly. Therefore, to maintain the government management and economic development needs, the funds invested in environmental protection are limited. Environmental pollution has obvious cross-regional characteristics in China, such as water pollution and air pollution. Therefore, in the actual environmental pollution control, local government environmental governance results are likely to be shared by neighboring areas, which further lessens the marginal benefits of local government environmental governance and makes regional governments less motivated to carry out environmental preservation.

At the same time, local governments cannot regulate their mechanisms, and individuals have lost their environmental rights. The failure of public policy also makes local governments unmotivated to protect the environment. There will be information concealment between the upper and lower levels in China’s administrative system. It will not have an excellent direct understanding of the local information. In reality, in the process of approval and construction, the relevant government departments claim that they meet the environmental standards. After an environmental pollution accident occurs, the public knows that the polluting enterprises do not meet the national environmental standards. In recent years, the blood lead incidents and mercury pollution incidents in Shandong, Shaanxi, and other places all show government and business collusion of local governments in attracting investment.

3.1.3. Agency Issues of Local Governments

China’s “directive policies” on environmental protection are mainly made by the central government, but the responsibility for spending is more often borne by local governments [21]. Due to information asymmetry and lack of social environmental protection concept, in order to maximize their own economic and political interests, grassroots governments distort the structure of financial expenditure to affect the industrial structure and realize the trade-off between economic growth and environmental protection services. Based on analyzing the differences between the expected goals of the central and local governments and the reality of environmental governance, this paper further analyzes the principal and agent problems of local governments in environmental governance under the condition of financial decentralization.

In China’s practice of environmental protection and pollution control, local governments are responsible for the implementation and supervision of environmental protection and pollution control policies formulated by the central government. However, the “upward responsibility” political system implemented in China and the local government governance mode combined with fiscal decentralization make the development of local economy bring economic incentives to local officials [22]. The model in this paper is based on Holmstrom and Milgrom model and has been improved and expanded in several aspects (combined with the reality of China) [23]. This paper sets up the central government as the principal, and the local government as the agent. The two tasks given in the goal are economic development and environmental protection. This article is presented in G and W, respectively. Accordingly, the local government’s expected output is $\mu =\left(u_g,u_w\right)$. To achieve this goal, local governments need to invest a lot of money and resources. Additionally, this study defines the efforts of local governments to achieve the above goals as $\mathrm{B}\left(b_g,b_w\right)$, corresponding to $C\left(G,W\right)$, in which u, B, C are greater than 0, and the first order can be guided. Finally, under the background of tax-sharing system, the payment scale of the central government to local governments is $\mathrm{P}\left(u_g,u_w\right)$, and the linear incentive function is $P\left(u_g,u_w\right)=\alpha +{\beta }^T\mu$, ${\beta }^T=\left({\beta }_g,{\beta }_w\right)$ which indicates the efforts of local governments in the two areas of economic construction and environmental protection. Assuming a linear relationship between the total local government output and the degree of effort, where ${\mu }_g=b_g+{\epsilon }_g, {\mu }_w=b_w+{\epsilon }_w$.

Assuming that local governments have a constant absolute risk avoidance bias $\mathsf{\rho }$, this paper sets the risk cost at $50\%\mathsf{\rho } ({\beta }_g^2{\sigma }_g^2+{\beta }_w^2{\sigma }_w^2)$.

Under these conditions, Local Government Participation Constraints (IR) and Incentive Constraints (IC) are indicated as follows:

$$\{\begin{array}{c}\left(IR\right)\alpha +{\beta }^T\mu -50\%\mathsf{\rho }({\beta }_g^2{\sigma }_g^2+{\beta }_w^2{\sigma }_w^2)-C\left(u_g,u_w\right)>P\\ \left(IC\right)\left(b_g,b_w\right)\in \arg maxP\left(u_g,u_w\right)-50\%\mathsf{\rho }({\beta }_g^2{\sigma }_g^2+{\beta }_w^2{\sigma }_w^2)-C\left(u_g,u_w\right)\end{array}$$

(3)

In the first derivative, the CE of the central government is 0, which can be obtained:

$$\{\begin{array}{l}\frac{\partial \left(CE\right)}{\partial \left({\mu }_g\right)}={\beta }_g-\frac{\partial C}{\partial \left({\mu }_g\right)}=0\\ \frac{\partial \left(CE\right)}{\partial \left({\mu }_W\right)}={\beta }_W-\frac{\partial C}{\partial \left({\mu }_W\right)}=0\end{array}$$

(4)

$$\mathrm{Solution}:{\beta }_g=\frac{\partial C}{\partial \left({\mu }_g\right)}, {\beta }_W=\frac{\partial C}{\partial \left({\mu }_W\right)}$$

(5)

Convert (5) into IR and CR constraints:

$$\{\begin{array}{c}Max{\beta }_g+{\beta }_W-\left(\alpha +{\beta }^T\mu \right)\\ s.t.\alpha +{\beta }^T\mu -50\%\mathsf{\rho }({\beta }_g^2{\sigma }_g^2+{\beta }_w^2{\sigma }_w^2)-C\left(u_g,u_w\right)\ge P\\ {\beta }_g=\frac{\partial C}{\partial \left({\mu }_g\right)},{\beta }_W=\frac{\partial C}{\partial \left({\mu }_W\right)}\end{array}$$

(6)

Convert the above formula to:

$$Max{\beta }_g+{\beta }_W-P-50\%\mathsf{\rho }({\beta }_g^2{\sigma }_g^2+{\beta }_w^2{\sigma }_w^2)-\mathrm{C}\left(u_g,u_w\right)$$

(7)

Further guidance:

$$\{\begin{array}{l}1-{\beta }_g-{\beta }_g\rho {\sigma }_g^2\frac{{\partial }^{2}C}{\partial {\mu }_g{}^2}-{\beta }_w\rho {\sigma }_w^2\frac{{\partial }^{2}C}{\partial {\mu }_g{\mu }_w}=0\\ 1-{\beta }_w-{\beta }_w\rho {\sigma }_w^2\frac{{\partial }^{2}C}{\partial {\mu }_w{}^2}-{\beta }_g\rho {\sigma }_g^2\frac{{\partial }^{2}C}{\partial {\mu }_g{\mu }_w}=0\end{array}$$

(8)

get:

$${\beta }_g=\frac{1-{\beta }_w\rho {\sigma }_w^2\frac{{\partial }^{2}C}{\partial {\mu }_g{\mu }_w}}{1+\rho {\sigma }_g^2\frac{{\partial }^{2}C}{\partial {\mu }_g{}^2}}$$

(9)

$${\beta }_w=\frac{1-{\beta }_g\rho {\sigma }_g^2\frac{{\partial }^{2}C}{\partial {\mu }_g{\mu }_w}}{1+\rho {\sigma }_w^2\frac{{\partial }^{2}C}{\partial {\mu }_w{}^2}}$$

(10)

$C_{gg}=\frac{{\partial }^{2}C}{\partial {\mu }_g{}^2},C_{gw}=\frac{{\partial }^{2}C}{\partial {\mu }_g{\mu }_w},C_{ww}=\frac{{\partial }^{2}C}{\partial {\mu }_w{}^2},C_{gw}>0.$ In addition, further simplify the upper formula. $C_{gw}$ represents the relationship between economic growth and environmental protection, if $C_{gw}>0$ means that the local government conflicts with the economic development and environmental protection tasks assigned by the central government; if $C_{gw}=0$, they are independent of each other, and if $C_{gw}<0$, the local government does not influence each other with the economic development and environmental protection tasks assigned by the central government. Further obtained:

$${\beta }_g=\frac{1-\frac{C_{gg}}{C_{gw}}+\frac{1}{\rho {\sigma }_w^2{C}_{ww}}}{\frac{1}{\rho {\sigma }_w^2{C}_{ww}}+\frac{{\sigma }_g^2{C}_{gg}}{{\sigma }_w^2{C}_{ww}}+1+\rho {\sigma }_g^2\left(C_{gg}-\frac{C_{gw}^2}{C_{ww}}\right)}$$

(11)

At the same time, because $C_w^2$ is a number going to infinity in the model, it can be further simplified:

$${\beta }_g=\frac{1-\frac{C_{gg}}{C_{gw}}}{1+\rho {\sigma }_g^2(C_{gg}-\frac{C_{gw}^2}{C_{ww}})}$$

(12)

This study states that under the current administrative system in China, the risk aversion degree of local governments is relatively stable in the short term, while the conflict between China’s economic development and environmental protection is still a state, so $C_{gw}>0$. When the economy and the environment are more uncoordinated, the corresponding ${\beta }_g$ will be greater. Additionally, the local economic construction value will be higher. Under multiple incentives, when the principal’s incentive for one goal is too strong, it will cause the agent to ignore the realization of the other goals. Additionally, the actual situation of our country, although the economic development and environmental governance belong to the goal of the central government, but compared with the environmental pollution control, the need to develop local economy is more urgent, and local government in the process of economic development by more incentive, thus leading to distribution on environmental protection weight ${\beta }_w$ will reduce accordingly, to reduce local government environmental protection functions [24].

In addition, under the principal–agent model, the attention of local governments to economic construction ${\beta }_g$ will also be affected by its possible degree ${\sigma }_g$. The derivation consequences of the previous paper indicate that the allocation weight of economic development goals ${\beta }_g$ is ${\sigma }_g$ a reduction function. The higher the observable degree of local government economic development completion, the greater the degree of effort. Under the current governance model of China, the GDP assessment system can be more convenient for assessing the performance of local officials, so it directly raises the weight of local allocation on economic development ${\beta }_g$. In contrast, due to the complex and changeable environmental protection tasks and imperfect observability, it is not easy to assess ecological protection results effectively. Local governments will allocate less weight to ecological protection accordingly ${\beta }_w$. A consequence of this is that local governments will reduce the scale or proportion of public welfare investment in environmental protection, medical care, and education for political achievements and tilt their limited financial resources to the economic field under the financial decentralization system, environmental severe pollution problems.

4. Empirical Analysis

4.1. Study Protocol Design

4.1.1. The Research Hypothesis Is Proposed

Based on the previous literature review and theoretical analysis, the finding is that local governments often invest limited resources in economic construction projects under the current financial decentralization system in China to obtain higher performance evaluation. Under the existing financial decentralization system, the central government collects most of the economic power, and the local government retains a part of the financial power for flexible allocation. To strengthen the control of the local governments, the central government will command the local governments through the distribution of financial management. Nevertheless, local governments adhere to economic construction as the core and vigorously promote economic development due to political performance assessment. Economic growth requires industrialization, foreign investment and infrastructure, and the higher the degree of local decentralization, the larger the scale of funds that local governments can spend on economic investment. At present, China’s economy is still on a relatively fast-growing trend, and on the whole, it is changing from extensive economic growth to an intensive economic growth model. Of course, the central and western regions are still dominated by a comprehensive economic growth model. In this context, to develop the financial demands, local governments often ignore the protection of the environment, which may bring serious pollution situations. Combined with the existing research results of scholars, this article proposes Hypothesis 1.

Hypothesis 1.

Financial decentralization will expand environmental pollution.

Foreign countries have a good awareness of environmental protection, the central enterprises have a high degree of environmental pollution control, and the domestic standards are lower than those of foreign countries, therefore attracting significant foreign capital investment. Thus, local governments utilize three ways to compete. These three types are: investment expenditure, taxation, and environmental regulation. Behind these three kinds of competition is to attract foreign investment. The fundamental reason for competition is political status. Local governments forced foreign capital and blindly implemented various policies to promote political promotion. Finally, the environmental problems have become more and more serious. So, it is inevitable that China will wear a “pollution shelter” hat. The reason for the emergence of pollution shelters is that, on the one hand, there are a large number of highly polluting enterprises investing in China. Due to the environmental regulations of the parent country, they have to transfer some industries with high pollution and high energy consumption to other regions. This behavior is also known as finding “environmental shelters” [25]. On the other hand, to promote regional economic development, the local governments have to lower environmental monitoring standards, create convenience for foreign enterprises to settle in, and further increase the environmental pollution in the region. In terms of financial structure, local governments cannot use funds freely, and to “free up” more funds for economic construction, they have to reduce the proportion of public financial expenditure on environmental protection, health, education, etc. Therefore, we propose Hypotheses 2 and 3.

Hypothesis 2.

Financial decentralization leads to intensified local competition, lowers the threshold of environmental protection, and expands the scale of pollution.

Hypothesis 3.

Financial decentralization will reduce the scale of public welfare fiscal expenditure, which is not conducive to curbing environmental pollution.

4.1.2. Variable Setting and Data Source

After reviewing the literature content and various materials, the author utilized the most appropriate one of the relevant models, namely the variable selection method. This article divides the variables into two categories. One is the explanatory variable, and the other is the explained variable. Among them, the degree of environmental pollution is the explained variable. Many explanatory variables include financial decentralization, foreign direct investment, per capita GDP, industrial structure, population density, etc. To build the model with the above data:

• Environmental pollution degree (pollute).

The author will use the three industrial wastes to measure them in the article. Additionally, that is the measurement of pollution. The three main primary pollutants are: industrial waste gas, wastewater and solid waste. Through the data analysis of the three factors, China’s local pollution index can be fully displayed. It is worth noting that there is no model for the overall weighted analysis of the three industrial wastes, and the author only uses a simple algebraic sum. The obtained new value and GDP are calculated as an indicator of environmental pollution.

2.

Financial decentralization (LFIS).

This metric is the core variable in the article. In general, the higher the target, the greater the local control over funds, and the more they will follow the guiding direction of the central government. The central government’s incentive program is directed in its direction. Sigman et al. [26] has linked environmental pollution to this indicator. There are many studies on financial decentralization, which many people have already studied. In this paper, the author thinks this index can be reduced: financial decentralization = fiscal revenue of provincial and municipal levels/national fiscal revenue. The index is all in the per-capita budget.

3.

Level of Foreign direct investment (FDI).

In today’s political performance assessment, foreign investment accounts for a considerable weight. So local governments attach great importance to this item. Furthermore, that is the core reason for the competition for local governments. External capital has a significant role in promoting local governments, which leads to local governments competing with each other. Nevertheless, at the same time, foreign capital is also a critical level of local economic development, but this indicates that its link with pollution is still very arbitrary, the writer also requires to explore its utility further.

4.

GDP per capita (lnpgdp).

Per capita GDP is the most authoritative indicator reflecting each province’s level of economic development. In this study, utilizing 2010 as the base period, CPI-adjusted actual GDP per capita GDP was used to measure the regional economic development level and square the per capita GDP variables to verify the Kuznets relationship between economic development level and environmental pollution and log the per capita GDP variables.

5.

Scientific and Technological Innovation (innovation).

In the existing environmental Kuznets curve research, the technical effect is considered an important reason for the inflexion point of the Kuznets curve. In general, technological progress will improve the efficiency of using resources and reduce energy consumption, thus inhibiting environmental pollution. Consequently, this paper uses per capita patent data as a proxy variable to measure scientific and technological innovation and logarithmically.

Additionally, this article also takes the number of vehicles (People) (Car), the proportion of financial expenditure on science and health (FIS), and industrial structure (Two) as control variables to further control the influence of economic and social factors on environmental pollution. In particular, this paper includes the degree of fiscal decentralization and the investment of foreign enterprises to explore how FDI affects places in it, which increases the weight of foreign investment on the environmental impact. At the same time, this paper will also explore the function mechanism of financial decentralization on environmental pollution by adding the variables of financial decentralization and scientific and technological innovation and the proportion of investment in science, education, culture and health. The definition and representation of each variable are illustrated in

Table 1. Definition, representation, and data sources of the variables.

Variable Name	Definition	Representation	Data Sources
Environmental pollution level	Three industrial waste emissions per ten thousand yuan of GDP (logarithmic)	pollute	The Urban Statistical Yearbook
Degree of fiscal decentralization	The per capita budget of the city level/the per capita national per capita budget	LFIS	The Urban Statistical Yearbook
Economic development level	GDP per capita (logarithmic)	lnpgdp	The Urban Statistical Yearbook
foreign direct investment	Scale of foreign direct investment	FDI	The Urban Statistical Yearbook
technological innovation	Number of patents per person (logarithmic)	innovation	Reference to Kou Zonglai and Liu Xueyue (2017)
density of population	Population per square kilometer (logarithmic)	People	The Urban Statistical Yearbook
Number of cars per capita	Number of cars per 10,000 persons (logarithmic)	Car	WIND data base
structure of fiscal expenditure	Scale of financial expenditure of science, education, culture and health	FIS	The Urban Statistical Yearbook
industrial structure	Value added of the secondary industry/GDP	Two	The Urban Statistical Yearbook

.

Descriptive statistical results of the main variables used in the model construction of this analysis are presented in

Table 2. Descriptive statistics of the variables.

Variable	Obs	Mean	Std.Dev.	Min	Max
pollute	3069	3.546	0.475	2.132	4.379
LFIS	3069	0.598	0.138	0.234	1.954
lnpgdp	3069	10.275	0.79	7.922	13.135
lnpgdp2	3069	106.21	16.454	62.765	172.525
FDI	3069	9.533	2.447	0.000	14.941
innovation	3069	1.048	1.127	0.000	6.968
People	3069	5.755	0.897	1.74	7.882
Car	3069	5.423	0.726	3.619	7.372
FIS	3069	0.199	0.045	0.016	0.387
Two	3069	0.436	0.085	0.213	0.599

, which also examines the differences in relevant indicators in 279 cities nationwide over 11 years between 2010 and 2020.

4.1.3. Model Construction

In this article, the data from 279 cities from 2010 to 2020. Due to the short panel, the autocorrelation between data may be low. To eliminate the effects of individual fixation and time effects. Based on Xue et al. [27], this paper adopts fixed effects model is used as a benchmark model to investigate the impact of fiscal decentralization on environmental pollution. The model is as follows:

$$\mathrm{pollute}=\mathsf{\alpha }+{\tau }_i\mathrm{LFIS}+{\beta }_i{\mathrm{X}}_i+\mathsf{\mu }+\mathsf{\lambda }+\mathsf{\epsilon }$$

(13)

In Equation (13), the regional environmental pollution variables are indicated by $\mathrm{pollute}$. ${\tau }_i$ represents the correlation between financial decentralization and environmental pollution; $\beta$ the fitting parameter of each control variable; $\mathsf{\mu }$ individual fixed effect; $\mathsf{\lambda }$ fixed time effect; $\mathsf{\alpha }$ intercept term; and $\mathsf{\epsilon }$ error item.

4.2. Empirical Process and Results

4.2.1. Positive Results at the National Level

Table 3. Analysis of the baseline regression results.

	OLS	RE	FE
LFIS	0.544 ***	0.077 **	0.074 **
	(0.048)	(0.036)	(0.035)
FDI	0.041 ***	0.012 ***	0.006 ***
	(0.003)	(0.002)	(0.002)
Innovation	−0.037 ***	−0.028 ***	−0.038 ***
	(0.009)	(0.008)	(0.008)
lnpgdp	0.797 ***	0.416 ***	0.427 ***
	(0.146)	(0.100)	(0.098)
lnpgdp2	−0.043 ***	−0.017 ***	−0.018 ***
	(0.007)	(0.005)	(0.005)
People	0.383 ***	0.147 ***	−0.023
	(0.008)	(0.013)	(0.016)
FIS	−0.541 ***	−0.385 ***	−0.416 ***
	(0.139)	(0.081)	(0.078)
two	0.868 ***	0.306 ***	0.505 ***
	(0.106)	(0.118)	(0.122)
Car	−0.025 ***	−0.002	0.004
	(0.008)	(0.006)	(0.005)
Cow distance item	−2.258 ***	5.097 ***	6.029 ***
	(0.762)	(0.510)	(0.504)
Howsman’s test		18.273 ***
sample capacity	3069	3069	3069
R-squared	0.118	0.178	0.156

Note: **/*** indicate the significance test at the 5%, 10% has been passed, and t value in parentheses.

indicates the benchmark regression results of this paper, with environmental pollution as the explained variable. OLS regression, random effect regression and fixed effects regression are performed. The Housman test results demonstrate that the fixed-effect model should be used, so the fixed-effect model is used as the benchmark model.

The estimated coefficient of the financial decentralization (LFIS) of the core explanatory variable in this paper is 0.074. It has passed the significance test at the 5% level, which indicates that improving financial decentralization will increase environmental pollution, which is consistent with hypothesis 1 of this paper. From the regression results of the OLS model and random effect model, the estimated coefficient of financial decentralization variables is still significantly positive.

From the perspective of per capita GDP variables, the primary term coefficient of per capita GDP variables was −0.427, and the second term coefficient of per capita GDP variables was 0.018, which all passed the significance test at the 1% level. The calculated inflexion point was obtained when the P GDP variable value was 11.86, which means that when the per capita GDP reached 141,600 yuan, the inflexion point of economic development appeared. From the perspective of the statistical analysis of variables, the PGDP variable value range between 7.922 and 13.135, economic development level over the inflexion point city only concentrated in Beijing, Shanghai is represented by a few eastern coastal developed cities, so when the economic development and environmental pollution, local officials have great motivation to develop the local economy as the central goal and ignore environmental governance. Additionally, in the fixed model, there is a significant inverted U-shaped relationship between environmental pollution and local economic development, which is consistent with the existing research views of scholars.

From the perspective of FDI variables, the estimated parameter of FDI was 0.006 and passed the significance test at the 1% level, which implies that FDI has worsened the environment in the region, showing that foreign investment has prominent “pollution shelter” characteristics in China. This paper argues that the formation path of the “pollution shelter” effect in China is mainly as follows: Firstly, the hidden transfer of pollution-intensive industries. Compared with developed countries, China’s environmental standards are low, coupled with local governments using various competitive means to attract foreign investment, and improve the return of foreign investment in China. The advanced technology and equipment imported by foreign capital to China do not directly export pollutants. However, in the production process, the pollutants are stuck in the place of investment, and China does not consider the environmental pollution brought by its production when undertaking the international industrial transfer. Secondly, foreign investment in China’s industrial expansion of environmental resources. The pricing of environmental resources is too low or priceless, and foreign investment has wholly entered all kinds of the resource mining industry in China. Through industrial expansion, the ecological environment is affected in various ways. Thirdly, the pollution penetration of the industrial chain. Multinational enterprises control the global value chain, and Chinese enterprises are at the end of the world industrial chain in the international division of labor. Multinational enterprises control through industrial division of labor, implement industrial chain pollution penetration, and transfer pollution industries to China. Fourthly, foreign capital imports “foreign garbage” in China. Foreign-funded enterprises directly invest in building factories in China and directly or indirectly cause environmental pollution by legally importing “foreign garbage” as a production input. Fifthly, foreign capital will not take the initiative to use advanced clean production technology, so the transfer of pollution industries to China cannot significantly improve China’s environmental protection technology, resulting in difficulty improving environmental quality.

From the perspective of scientific and technological innovation, the estimated coefficient of Innovation variable is −0.038, and it has passed the significance test at the 1% level, which shows that technological progress is conducive to inhibiting environmental pollution, which is consistent with the existing research views of scholars and consistent with the previous description. Technological progress has brought about the improvement of resource utilization efficiency. With technological progress, some clean energy sources gradually replace fossil energy and reduce the pollution problem in economic development. From the perspective of other control variables, the population density variables did not pass the significance test in the fixed-effect model. In contrast, the population density variables were significantly positive in the OLS and the random-effect models, showing that increasing population density will expand the environmental pollution problem. From the perspective of the structural variable of fiscal expenditure, the estimation coefficient of FIS of the proportion of fiscal expenditure in science, education, culture and health is significantly negative. It has passed the significance test at the 1% level, which shows that increasing the proportion of fiscal expenditure on science and education is conducive to reducing environmental pollution. A possible explanation is that because environmental pollution is a typical public good, the market is not strongly motivated to actively control environmental pollution, and the government is still the main participant in environmental pollution control. The more significant the proportion of spending on science, education, culture, and health means that local governments pay more attention to science and technology, education, culture, and health, so the more substantial the motivation to control environmental pollution, conducive to reducing the environmental pollution. From the perspective of industrial structure variables, the more significant the proportion of the secondary industry, the more serious the environmental pollution, which shows that the development of China’s secondary industry is still an extensive development mode and does not pay much attention to the environmental value.

4.2.2. Mechanism Inspection

As stated in this study, the growth in local decentralization will give local officials more financial resources to manage behavior. In order to increase the growth of the demand economy, on the one hand, local governments will promote the rapid economic growth within the jurisdiction by attracting foreign capital, and scholars also consider this phenomenon as a horizontal competition among governments. At the same time, within the limited scope of local financial resources, local officials may invest more money in economic development to develop the local economy, accordingly, squeezing spending on science, education, culture, and health. Therefore, this paper empirically examines the mechanism of local decentralization and environmental pollution by adding the form of interaction terms. See Equations (2) and (3) for details.

$$\mathrm{pollute}=\mathsf{\alpha }+{\tau }_i\mathrm{LFIS}+{\xi }_i\mathrm{LFIS}\times \mathrm{FDI}+{\beta }_i{\mathrm{X}}_i+\mathsf{\mu }+\mathsf{\lambda }+\mathsf{\epsilon }$$

(14)

$$\mathrm{pollute}=\mathsf{\alpha }+{\tau }_i\mathrm{LFIS}+{\theta }_i\mathrm{LFIS}\ast \mathrm{FIS}+{\beta }_i{\mathrm{X}}_i+\mathsf{\mu }+\mathsf{\lambda }+\mathsf{\epsilon }$$

(15)

The marginal utility method is adopted to differentiate Formulas (14) and (15) for fiscal decentralization $\mathrm{LFIS}$, it can be found that the first order guide ${\tau }_i$ and the ${\xi }_i\mathrm{FDI},{\theta }_i\mathrm{FIS}$. When ${\xi }_i$ significantly positive, the influence of financial decentralization on environmental pollution will expand with the improvement of foreign capital utilization level. When ${\theta }_i$ significantly negative, the influence of financial decentralization on environmental pollution will expand with the proportion of science and education spending.

Table 4. Test of the mechanism of action.

	Pollute	Pollute	Pollute	F IS	FDI
LFIS	0.074 **	0.045 ***	0.549 ***	−0.052 ***	0.892 ***
	(0.035)	(0.089)	(0.088)	(0.008)	(0.314)
FDI	0.006 ***	0.026 ***	0.006 ***	0.001 **
	(0.002)	(0.006)	(0.002)	(0.001)
Innovation	−0.038 ***	−0.048 ***	−0.035 ***	0.012 ***	−0.111
	(0.008)	(0.008)	(0.008)	(0.002)	(0.068)
lnpgdp	0.427 ***	0.399 ***	0.433 ***	−0.150 ***	−2.217 **
	(0.098)	(0.098)	(0.098)	(0.024)	(0.887)
lnpgdp2	−0.018 ***	−0.018 ***	−0.019 ***	0.007 ***	0.054
	(0.005)	(0.005)	(0.005)	(0.001)	(0.044)
People	−0.023	−0.022	−0.021	−0.004	−0.050
	(0.016)	(0.016)	(0.016)	(0.004)	(0.141)
FIS	−0.416 ***	−0.401 ***	1.038 ***		1.667 **
	(0.078)	(0.077)	(0.259)		(0.702)
two	0.505 ***	0.611 ***	0.482 ***	−0.012	−0.764
	(0.122)	(0.123)	(0.121)	(0.030)	(1.105)
Car	0.004	0.007	0.002	−0.000	−0.100 **
	(0.005)	(0.005)	(0.005)	(0.001)	(0.050)
LFIS × FDI		0.055 ***
		(0.009)
LFIS × FIS			−2.525 ***
			(0.430)
Cow distance item	6.029 ***	6.088 ***	5.791 ***	1.055 ***	−6.297
	(0.504)	(0.501)	(0.502)	(0.121)	(4.555)
sample capacity	3069	3069	3069	3069	3069
R–squared	0.156	0.167	0.167	0.073	0.156

Note: **/*** indicate the significance test at the 5%, 10% has been passed, and t value in parentheses.

shows the regression results of the mechanism of action presented in this paper. Taking the expenditure on science, education, culture and health and the level of foreign capital utilization as the explanatory variables, it can be found that financial decentralization will reduce the proportion of expenditure on science, education, culture, and health and increase the level of foreign capital utilization. From the perspective of the horizontal competition mechanism of local governments, the higher the decentralization of local finance, the more funds used to attract foreign capital to occupy a favorable position in the competition of local governments. Taking environmental pollution as the explained variable, it can be found that the financial decentralization coefficient is reduced from 0.074 to 0.045, while the estimated coefficient of LFIS × FDI is 0.055 and has passed the 1% significance test, which shows that fiscal decentralization can increase environmental pollution through channels to strengthen the competition between local governments. From the perspective of fiscal expenditure structure proportion, after adding the fiscal expenditure structure and fiscal decentralization interaction, the estimated coefficient rose from 0.074 to 0.549, and the fiscal expenditure structure and fiscal decentralization interaction estimated coefficient of −2.525 and are at 1% level passed the significance test, it shows that financial decentralization can influence the fiscal expenditure structure and affect the environmental pollution. Specifically, the expansion of fiscal decentralization will reduce local governments’ fiscal spending on science, education, culture, and health and use more funds for economic construction, thus increasing environmental pollution.

4.3. Robustness Test

4.3.1. Variable Replacement Test

In the empirical benchmark results of this paper, there is a significant positive correlation between financial decentralization and environmental pollution. In the current environment of local economic development in China, fiscal decentralization will stimulate the motivation of local government officials to develop the economy and ignore the pollution problem in the process of economic development, which is consistent with the theoretical derivation of this paper. Furthermore, a series of robustness tests are conducted to support this conclusion in this section.

First, based on substitution variables, this article changes the measure of fiscal decentralization to the fiscal self-sufficiency index, taking the proportion of local fiscal general expenditure to local fiscal revenue as the proxy variable of fiscal decentralization. Column 1 in

Table 5. Robustness checklist.

	Pollute	Lnpwater	Lnpso2	Lnpyan
LFIS	0.017 ***	0.689 ***	0.937 ***	0.361 *
	(0.004)	(0.116)	(0.157)	(0.197)
FDI	0.006 ***	0.013 *	0.004	−0.016
	(0.002)	(0.007)	(0.009)	(0.012)
Innovation	−0.030 ***	−0.122 ***	−0.193 ***	−0.007
	(0.007)	(0.025)	(0.034)	(0.043)
lnpgdp	0.369 ***	0.185	1.449 ***	−0.679
	(0.095)	(0.326)	(0.444)	(0.556)
lnpgdp2	−0.016 ***	−0.003	−0.074 ***	0.055 **
	(0.005)	(0.016)	(0.022)	(0.028)
People	−0.023	0.049	−0.060	−0.185 **
	(0.016)	(0.052)	(0.071)	(0.088)
FIS	−0.440 ***	0.012	0.576	−0.015
	(0.078)	(0.258)	(0.352)	(0.440)
two	0.568 ***	2.473 ***	0.350 **	2.207 ***
	(0.120)	(0.406)	(0.153)	(0.692)
Car	0.006	0.048 ***	0.128 ***	0.102 ***
	(0.006)	(0.018)	(0.025)	(0.031)
Cow distance item	5.769 ***	1.133	−0.731	6.744 **
	(0.495)	(1.673)	(2.279)	(2.854)
sample capacity	3069	3069	3069	3069
R–squared	0.155	0.086	0.186	0.045

Note: */**/*** indicate the significance test at the 1%, 5%, 10% has been passed, and t value in parentheses.

shows the regression results of replacement financial decentralization variables, which show a significant positive correlation between financial decentralization and environmental pollution after the replacement, which is consistent with the benchmark regression results in this paper. The 10 main variables are consistent with the benchmark regression model, reflecting the robustness of the empirical results.

Further, the industrial wastewater, waste gas, and waste are returned separately, and the results are shown in columns 2–4 of Table 5. All financial decentralization variables were significantly positive and at least passed the significance test at the 10% level. The OFDI variables only showed a significant positive correlation with the water pollution discharge among the other variables. Technological progress indicates a significant negative correlation between industrial wastewater discharge and industrial waste gas. Environmental Kuznets curve relationship does not necessarily exist between economic development and industrial wastes. In the empirical regression results of this paper, there is only a significant environmental Kuznets curve relationship between air pollution and economic development level. It is worth noting that the secondary industry structure, car ownership and industrial wastes all demonstrate a significant positive correlation.

4.3.2. Heterogeneity Analysis

Taking the pollution level as the standard, the sample is divided into high pollution areas and low pollution areas according to the median and returned, respectively. The specific results are shown in

Table 6. Heterogeneity analysis.

	High Pollution	Low Pollution	Provincial Capital	Non–Provincial Capital
LFIS	0.093 **	0.060	−0.014	0.108 **
	(0.045)	(0.052)	(0.051)	(0.050)
FDI	0.005 **	0.008	0.010 *	0.003 **
	(0.002)	(0.005)	(0.006)	(0.001)
Innovation	−0.075 ***	−0.023 **	−0.042 ***	−0.039 ***
	(0.012)	(0.011)	(0.013)	(0.010)
lnpgdp	−0.348 ***	−0.564 ***	−0.457 ***	−0.352 **
	(0.134)	(0.150)	(0.149)	(0.150)
lnpgdp2	0.015 **	0.024 ***	0.019 **	0.015 **
	(0.007)	(0.007)	(0.008)	(0.007)
People	−0.035	−0.012	−0.005	−0.042 *
	(0.026)	(0.019)	(0.021)	(0.023)
FIS	−0.497 ***	−0.349 ***	−0.349 ***	−0.509 ***
	(0.114)	(0.107)	(0.111)	(0.109)
two	0.665 ***	0.233	0.231	0.797 ***
	(0.158)	(0.221)	(0.155)	(0.211)
Car	0.003	0.005	−0.001	0.017*
	(0.008)	(0.008)	(0.006)	(0.010)
Cow distance item	5.234 ***	7.120 ***	6.154 ***	5.724 ***
	(0.682)	(0.785)	(0.744)	(0.785)
sample capacity	1540	1529	330	2739
R–squared	0.189	0.143	0.191	0.187

Note: */**/*** indicate the significance test at the 1%, 5%, 10% has been passed, and t value in parentheses.

. It can be found that in high pollution areas, financial decentralization will increase the degree of pollution level, while financial decentralization in low pollution areas will not impact pollution. However, FDI and industrial structure variables are also significantly positive in high pollution areas and not significant in low pollution areas. In this paper, we believe that the differences in the impact of fiscal decentralization in different polluted areas can still be analyzed from the orientation of local economic development goals. Although the pollution prevention and control situation in high pollution areas is relatively more complicated, it should be noted that the local government’s lack of attention to environmental pollution leads to increasingly serious environmental pollution problems. In the pursuit of faster economic growth, the improvement of the local government fiscal decentralization will bring local government game ability to attract more foreign investment. At the same time, in order to promote employment and boost the economy, local governments may strengthen the investment in regional secondary industry. This series of economic growth as the only goal-oriented behavior eventually led to high pollution.

In addition, this paper divides the sample cities into provincial and non-provincial cities. The basis of this division is as follows: this paper argues that although local governments are facing the pressure of economic development, the pressure varies between different levels. Provincial capital cities are the pillar of regional economic development, the growth pole of the province’s economic growth, and the object of national critical investigation. Compared with the inspection faced by prefecture-level cities, the national investigation of provincial capitals includes the growth of economic indicators and includes a large number of environmental protection and sustainability indicators. Therefore, provincial capitals will not necessarily use environmental pollution in exchange for economic growth. Table 6 also reports the regression results of provincial capitals and non-provincial capital cities, which shows that financial decentralization is more likely to bring environmental pollution problems in non-provincial capital cities. This is consistent with the previous analysis, where prefecture-level city governments do not consider environmental protection when facing the pressure of economic development, while provincial capitals cannot sacrifice environmental indicators to drive economic growth due to national restrictions. At the same time, it also explains that financial decentralization brings the contradiction between the national, provincial and prefecture-level governments in facing economic development and environmental protection. The national-level hopes to protect the environment, while local governments often ignore environmental protection for political achievements. Provincial capitals facing strong supervision of the state may reduce the environmental damage to economic development, while prefecture-level city governments have faceless supervision, so local decentralization brings greater convenience for local governments to intervene in the economy. From the comparison of other variables, the structure variables of the secondary industry will bring more serious environmental pollution problems in non-provincial cities, while there is no inevitable connection with the environmental pollution in provincial cities.

5. Conclusions and Recommendations

5.1. Study Conclusion

The existing fiscal decentralization system is essential to China’s economic system reform and vital in stimulating local economic development. However, under the existing political incentive system, the greater the financial autonomy of local governments to obtain a higher performance evaluation, the more monetary funds will be invested in economic construction, and the protection of the local environment will be ignored. In some way, China’s existing environmental pollution problem results from the distortion of Chinese fiscal decentralization and local government incentives.

By explaining the connotation and characteristics of fiscal decentralization in China and comparing the differences between domestic fiscal decentralization practice and western fiscal decentralization theory, this paper aims to clarify the possible influence of fiscal decentralization on environmental pollution. At the same time, this paper analyzes the behavior of local government under the decentralization system, expounds on the differences between the central and local government expected goals and not dynamic under the existing incentive mechanism, at the same time based on the principal–agent model to further clarify the local government and the central government in the contradiction between economic development and environmental governance. This article analyzes the local government behavior and environmental pollution from local government competition and fiscal expenditure. Based on literature research, this paper empirically examines the environmental finance of cities from 2010 to 2020 and establishes a panel data model. After careful analysis of the data, we draw the following conclusions.

China’s fiscal decentralization increases pollution emissions, but there are differences in different regions.

Financial decentralization will aggravate the competition of local governments. The higher the level of financial decentralization, the higher the level of foreign investment utilization, and the proportion of public welfare financial expenditure on science, education, culture and health will be relatively reduced. However, foreign investment has prominent “pollution shelter” characteristics in China, so enhancing financial decentralization levels increases environmental pollution.

The proportion of secondary industry structure will increase environmental pollution, while technological progress will reduce environmental pollution.

5.2. Policy Suggestions on Strengthening Environmental Governance under the Background of Financial Decentralization

5.2.1. Constructing the Fiscal Policy of Environmental Governance

We will establish a particular environmental public finance system. At present, in the level of environmental fiscal revenue, the measures can be taken, including environmental protection tax, administrative punishment and other measures. At the level of environmental financial expenditure, the government proposed establishing an environmental budget system, preparing for a rainy day, and budgeting environmental expenditure funds arrangement to avoid the phenomenon of seizing and private occupation of funds. Moreover, the government needs to consider sustainable development to ensure that the environmental budget system improves and spending continues to grow with the economy.

5.2.2. Improving the Existing Fiscal Decentralization System

We will clarify the administrative and financial powers of the central and local governments in environmental governance. The current environmental governance mechanism should further clarify that the central government and the local government are in environmental governance. Power and financial power in the management. In terms of administrative power, the central government should further play the role of macro-control to formulate national environmental protection strategies, policies, laws and regulations, increase the frequency of environmental protection supervision, and at the same time establish special environmental protection funds to encourage local environmental protection administrative acts. It will be apparent that both central and local governments are responsible for environmental governance and that the authorities will sort out the relationship between the market and the government. The government will no longer play a full role but will let the market play its original regulatory role. When solving environmental problems, such as the construction of environmental governance infrastructure, the government should encourage the market to solve the problem.

5.2.3. Improving Local Incentive Mechanism

We apply to the proposal that government departments focus on environmental governance while focusing on economic growth and environmental improvement. Enterprises and the government should add a more scientific and reasonable assessment mechanism in the assessment, and the assessment system should also combine environmental governance and economic development to calculate “green GDP” in the system.

In order to implement the green GDP assessment work, the government needs to take into account the negative externalities of the economy and the positive externalities of environmental improvement and implement the corresponding policies to remove the cost of environmental governance in the original GDP value.

5.2.4. Changing the Pattern of Economic Development

The secondary industry is the primary source of environmental pollution in China and the secondary industry. The larger the proportion, the larger the scale of environmental pollution. Therefore, local governments should seize the opportunity of industrial transfer, change the extensive economic development mode, drive the upgrading of industrial structure with technological progress, and gradually develop the dependence of economic development from the high-pollution secondary industry to a cleaner tertiary industry. Moreover, we need to go even further to enhance our environmental awareness. In introducing foreign investment, the threshold of local environmental protection should be raised, and the requirements of environmental protection measurements and policies should only meet the requirements to prevent China from becoming a “pollution refuge” for foreign investment.