Has Central Environmental Protection Inspection Promoted High-Quality Economic Development?—A Case Study from China

Abstract

Environmental regulatory policies are crucial to the high-quality development of China’s economy. Can central environmental protection inspection, another major innovation in China’s environmental regulation policy, significantly contribute to high-quality economic development? To address this question, this paper is the first time to intensely discuss the relationship between central environmental protection supervision and the high-quality development of China’s urban economy. The research level is expanded from the previous micro-enterprise level to the macro level. Starting from the city, the main body of green innovation, and using the propensity score matching and difference-in-differences methods, we test the impact of central environmental protection inspection on high-quality economic development. A preassessment at the regional level reveals that the green total factor productivity of the inspected cities is significantly higher than that of the non-inspected cities, and parallel-trend tests and placebo tests also support the result. Subsequently, the persistence of the policy impact is further analyzed, and the results show a significant impact only in the year of policy implementation. Finally, the heterogeneity analysis shows that central environmental protection inspection can significantly promote high-quality economic development in cities in the eastern region, but has a significant inhibitory effect on the western region.

1. Introduction

With the rapid development of China‘s economy, resource waste, inefficient energy use, environmental pollution and other problems have emerged. In 2013, the Asian Development Bank made clear in its environmental analysis report [1] on air quality standards formulated by the WHO that less than 1% of China’s 500 large- and medium-sized cities met air quality standards in 2012. Furthermore, in the Global Environmental Performance Index rankings published in 2018, China ranked just 120th in terms of the overall score. Accelerating the transformation of economic development modes is imminent. China urgently needs to develop the green economy vigorously to solve the fundamental contradiction between environmental deterioration and economic growth. According to the Porter hypothesis, a moderate level of environmental regulation can stimulate technological innovation and achieve a win–win situation between economic growth and environmental protection [2]. Therefore, exploring the impact of environmental regulation on the high-quality development of urban economies is the most essential part of achieving high-quality economic development in China [3]. However, the traditional “command-and-control” environmental regulation policies are generally plagued by “collusion between government and enterprises”, “incentive distortion”, “formalization” and data falsification. It has been documented that, in the context of financial and promotion incentives, local governments have neglected environmental protection in the pursuit of economic development and “government–enterprise collusion” has become a direct cause of poor environmental governance [4]. In addition, the “promotion tournament theory” suggests that the government is motivated by this behavior to develop the economy at the expense of environmental quality and that the secondary status of environmental protection in performance appraisal directly weakens the motivation of local governments to protect the environment, creating an “incentive distortion” [5]. Ghanem et al. point out that China’s environmental governance model tends to ignore the construction of long-term mechanisms and suffers from the defects of “cutting at one stroke” policy implementation and “A gust of wind” treatment effects [6]. These problems cause policy effects to deviate from expectations and even affect economic development [7].

In the current development dilemma, a policy with characteristics of Chinese environmental governance has emerged. In January 2016, central environmental protection inspection (CEPI) took the lead in Hebei, and in two years covered the country’s 31 provinces (autonomous regions and municipalities); on behalf of the Party Central Committee and the State Council and given the authority of environmental protection inspectors Karisma, emphasizing the work of environmental protection in terms of “party and government together” and “one post, two responsibilities”, scholarly studies have shown that CEPI has had significant and sustained effects in improving air quality [8]. However, in the context of increasing ecological constraints, supply-side institutions have become another voice in the conversation regarding the impact that CEPI can have on economic growth and will face a greater impact under environmental protection inspection. The Ningxia Economic and Information Commission has repeatedly reported to the autonomous regional government that environmental protection inspection and other work affect economic development. Whether or not environmental regulation can help high-quality economic development has become a hot topic for scholars. As another significant innovation in environmental regulation policy, can CEPI actually affect the quality of economic development? Does CEPI have regional heterogeneity with high-quality economic development? To address the above issues, we draw on the research idea of Yu Yongze et al. [9] to use green total factor productivity (GTFP) as a proxy for high-quality economic development and adopt an empirical approach to assess the policy effects of CEPI.

2. Policy Background and Literature Review

2.1. Policy Background

Since 2013, when General Secretary Xi Jinping proposed “deepening the reform of ecological civilization system”, China has been innovating environmental protection policies to address local governments’ environmental governance dilemmas and to facilitate high-quality economic development. CEPI, another major innovation in the “command-and-control” type of environmental regulation, is known as the most extensive environmental protection activity organized at the national level in China for its broad scope, strength, and targeting. Beginning on 4 January 2016, the Ministry of Ecology and Environment launched a particular inspection campaign for five batches of 31 provinces, as shown in Figure 1, which ended on 15 September 2017. The first round of inspectors and “look back” personnel received a total of more than 212,000 reports from the public, the number of accountable people exceeded 17,000, the closure of more than 200,000 chemical enterprises was initiated, and the number of chemical enterprises to be rectified reached nearly 1,000,000. CEPI differs from previous environmental regulatory policies with serious overtones of local protection, ensuring that enforcement is independent and influential. The reasons for this can be categorized into three points: (1) The central environmental protection inspection team is randomly dispatched by the State Council and is subordinate to it, ensuring its independent enforcement. (2) CEPI takes the route of listening to reports, having access to information, accepting interviews, conducting individual conversations and receiving reports to avoid the problem of information asymmetry effectively. (3) By opening up multiple paths for mass participation, CEPI specifically sets up a reporting channel to ensure that reports on environmental protection calls and letters are accepted as soon as possible. The public is actively involved in environmental governance under the supervision of the CEPI system, and there is a reasonable use of public opinion to monitor government pollution-governance behavior [10].

2.2. Literature Review

The research object of this paper is high-quality economic development. Compared with the traditional index system, more and more theoretical and empirical studies show that GTFP is an effective indicator in the measurement of high-quality economic development [11]. Total factor productivity (TFP) is used to measure the efficiency of resource utilization and development. GTFP is dedicated to a more comprehensive productivity measurement by considering energy and resource factors based on total factor productivity. Therefore, this part of the literature review focuses on two aspects: the literature on the path to high-quality economic development and the literature on CEPI.

2.2.1. A Review of Development Paths for High-Quality Economic Development

High-quality economic development emphasizes the simultaneous growth of the quality and quantity of economic growth, which is an important criterion to objectively evaluate the level of China’s economic development under the innovation drive. Measuring the level of economic quality development has also become a hot topic of academic research. Unlike the previous single indicator reflecting incomprehensiveness and limitations, more and more scholars have found that the GTFP can effectively measure economic quality development. Endogenous growth theory suggests that TFP is a powerful explanation for regional income and economic disparities. At the same time, GTFP takes environmental pollution as an unexpected output into the calculation model, which is the proper meaning of high-quality economic development. Academics are increasingly inclined to use it to explain economic disparities [12]. Early total factor productivity research focuses on technical efficiency, often ignoring the importance of the environment. With rapid economic development, environmental problems are becoming more and more serious, and people are beginning to devote more attention to environmental protection. More and more scholars put environmental factors (industrial wastes) into the model calculation. Drawing on international experience, Hu et al. [13] used a directional distance function model to measure the technical efficiency of Chinese provinces from 1999 to 2005. The early industrial field is the focus of green total factor research. For example, Tu et al. [14] found that instead of inhibiting industrial development, environmental regulation in China found that GTFP became a central driver of industrial growth and environmental management. Chen [15] estimated the TFP changes in the Chinese industry using a sub-industry production function and found that the Chinese industry has generally achieved an intensive growth transformation. However, some high-emission industries still need to improve energy-saving and emission-reduction technologies to achieve sustainable development. Chen [16] further found that GTFP taking environmental factors into account, was much lower than traditional TFP. It has also been found that a series of energy conservation policies significantly increased GTFP, suggesting that total factor productivity measured by setting aside environmental factors would seriously underestimate the environmental costs sacrificed for economic growth.

2.2.2. Literature Review on Central Environmental Protection Inspection

CEPI differs from traditional campaign-based environmental governance with authority and continuity. On the one hand, central environmental protection inspectors mobilize various resources within the system from top to bottom and vigorously promote environmental protection inspection work by the central authority. A campaign-type environmental governance mechanism is adopted to correct the failure of the existing conventional-type environmental governance mechanism [17]. On the other hand, central environmental protection inspection has a political and social interaction and continuous governance of the normalization of development trends is undertaken. In July 2019, the second round of CEPI was launched. It is expected to take four years to complete full coverage and “look back”, focusing on pollution prevention, control of the battle, and high-quality economic development.

In the existing literature on CEPI, academics focus on the pollution control effect of the policy. Wu and Hu used breakpoint regression to study the effect of CEPI on the air quality index, and the results showed inter-provincial differences in the impact of the policy [18]. Moreover, the effect of CEPI on air pollution is mainly reflected in PM10 and PM2.5 from the results of pollutants. Tu et al. explored the environmental and economic benefits of central environmental protection inspection from three perspectives: ecological, economic, and social benefits in Hebei Province using the breakpoint regression method. The results showed that it significantly improved the air quality in Hebei Province with long-term effects [19]. Wang et al. used prefecture-level cities as the study target. They found that pollutants in the inspected cities were significantly reduced compared to non-inspected cities, and the reduction was slightly weaker in the “look-back” period [20]. While the above-mentioned environmental governance effects were developed at the regional level, Wang and Fan evaluated the effectiveness of campaign-based enforcement in promoting corporate environmental actions from a micro-enterprise perspective using central environmental protection inspection as the study target [8]. The study found that the number of polluting enterprises decreased 48% after implementing CEPI. In addition, central environmental protection inspection significantly reduced industrial chemical oxygen demand (COD) by at least 46.5%.

A few kinds of literature have focused on the economic effects of central environmental protection inspection [21]. CEPI may have two effects on micro enterprises’ production and operation activities. One is to follow the cost effect, that is, strict environmental governance policies which increase the cost of pollution emission and the governance of enterprises in the city, thus increasing productive investment and innovative R&D investment and reducing green innovation capacity [22]. The second is the innovation compensation effect. To put it another way, by increasing investment in environmental protection, enterprises can enhance the level of green innovation and reduce the cost pressure brought by the action of the inspectors. Therefore, the innovation compensation cost effect can be stimulated, becoming more remarkable than cost effect, and thus improving the level of green innovation of enterprises in pollution-intensive industries. Finally, a win–win situation of simultaneous improvement in environmental and economic performance is likely to be achieved [23,24]. In addition, some scholars have explored the relationship between central environmental protection inspection and the share price of enterprises. Chen believes that, in the short term, central environmental protection inspection has a significant impact on the share prices of both types of enterprises as it changes investors’ expectations. In the long term, the share prices of environmental pollution-oriented companies are more affected by this policy, and central environmental protection inspection does not achieve the role of social capital reallocation [25]. Tian et al. used a sample of 270 listed companies in the heavy pollution industry in China and used the event study method to conclude that central environmental protection inspection has a negative impact on the share prices of heavy pollution companies [26]. Meanwhile, politically affiliated companies suffer less wealth loss during the ten-day activity window compared to non-politically affiliated companies, suggesting that political affiliation can mitigate the negative impact of CEPI on the stock value of heavily polluting companies.

From a comprehensive perspective, domestic and foreign scholars’ research on the policy impact of central environmental protection inspection presents an all-round and multi-level character. However, it can be improved from the following points: First, there is minimal research on the economic effects of central environmental protection inspection. Research mostly explores the effect of environmental management on environmental pollution and the mechanisms of this effect, and lacks exploration of the impact of the policy on high-quality economic development. Second, studies on the contribution of economic growth lack the use of macro-city-level data, and most use enterprise panel data. They never study the impact of central environmental protection inspection on the high-quality development of urban economies from the perspective of cities as the main body of green innovation.

Given this, this paper conducts an empirical study of the economic effects of the policy using the propensity score matching (PSM) and difference-in-differences (DID) methods. Additionally, a series of robustness tests are conducted for the first two inspected cities in the first round of central environmental protection inspection, using city-level data from 2013 to 2019 for prefecture-level cities in China. The main contributions of this thesis are summarized as follows: First, it enriches the research literature on central environmental protection inspection. CEPI is a significant party initiative on ecological and environmental protection in the new era. The “party and government share responsibility” and “one post, two responsibilities” require close cooperation between local party committees and governments. This paper empirically tests whether CEPI can achieve a win–win situation of simultaneous environmental and economic performance improvement. Second, for the first time, central environmental protection inspection is linked to the high-quality economic development of cities for an in-depth discussion. Expanding the research level from the previous micro-firm level to the macro level, the impact of central environmental inspection on high-quality economic development is examined from the perspective of the city as a subject of green innovation. This is a valuable addition to the literature on the inspection system and its governance effects, and provides empirical evidence for optimizing local environmental governance inspections and inspection practices in other areas. Third, the PSM and DID methods are used to identity the net effect of CEPI to effectively avoid the endogeneity problems arising from selection bias and the use of other measures, and heterogeneity analysis with a placebo test is conducted.

3. Empirical Model and Variable Description

3.1. Model Setting

3.1.1. Difference-in-Differences Model Setting

The DID idea is used to consider implementing a new policy as a natural experiment and to test the difference in the mean change in the experimental group affected by the policy shock compared to the control group to estimate the policy effect. Central environmental protection inspection provides a significant quasi-natural experimental opportunity. In this paper, DID is used to analyze the changes in GTFP in the inspected cities compared to the non-inspected cities to explore the impact of central environmental protection inspection on the high-quality economic development of cities.

$$Y_{it}={\delta }_0+{\delta }_{1}CEP{I}_{it}\times yea{r}_{it}+{\delta }_2{X}_{it}+{\mu }_i+{\lambda }_t+{\epsilon }_{it}$$

(1)

where the subscripts $i$ and $t$ are individual and time, $CEP{I}_{it}$ a represents the central environmental protection inspection policy dummy variable. Cities that were mentioned or had air pollution problems in the 2017 rectification and feedback program take the value of 1, or otherwise take the value of 0. In addition, $yea{r}_{it}$ is a time dummy variable that takes the value of 0 before 2017, and 1 for 2017 and beyond. $Y_{it}$ represents the dependent variable for city $i$ at time $t$, measured using green total factor productivity. The study collects control variables $X_{it}$, Including city size (Popu), economic development (RGDP), industrial structure (Indu), level of financial development (Fin) and trade openness (FDI). In addition, we use a two-way fixed-effect model: ${\mu }_i$ is individual fixed effects, ${\lambda }_t$ is time-fixed effects, and ${\epsilon }_{it}$ is errors.

3.1.2. GTFP Measurement

GTFP is the residual value after proposing factor inputs, and it introduces factors such as energy, resources, and undesired output based on total factor productivity. It increases the endogenous motivation for reducing energy and resource consumption and for energy reduction behavior, which reflects the new development concept. Undoubtedly, GTFP is an essential indicator for measuring high-quality economic development. For the measurement of GTFP, considering that when there is non-zero slack in inputs or outputs, traditional DEA models can ignore specific characteristics of inputs or outputs, resulting in the over- and underestimation of efficiency. Therefore, the non-radial and non-angle SBM (slacks-based measure) model proposed by Fare et al. [27] and Fukuyama et al. [28] is adopted to effectively solve the problem of output slack by introducing slack variables.

(1)

SBM directional distance function: In this paper, each prefecture-level city is considered as a production decision unit from which the optimal production frontier surface in each period for each of the 253 prefecture-level cities in China is constructed. Following the idea of Fare, assuming that each city uses $N$ factor inputs $x=(x_1,\dots ,x_n)\in R_n^{+}$. Obtain $L$ desired outputs $y=\left(y_1,\dots ,y_l\right)\in R_l^{+}$ and $M$ undesired outputs $b=\left(b_1,\dots ,b_m\right)\in R_m^{+}$. Assume that there are $t$ periods and $c$ decision units, where $c=1,2,\dots ,C$, $t=1,2,\dots ,T$, then the input and output values of city $c$ at time $t$ can be simply expressed as $\left(x_c^t,y_c^t,b_c^t\right)$. Based on the above assumptions, the directional distance function considering non-desired outputs is defined as follows.

$$\begin{array}{l}{S}^{*}(x_c^{t^{\prime }},y_c^{t^{\prime }},b_c^{t^{\prime }},g^x,g^y,g^b)=\max \left[\frac{1}{2N}{\displaystyle {\sum }_{n=1}^N\frac{s_n^x}{g_n^x}+\frac{1}{2L+2M}\left.({\displaystyle {\sum }_{l=1}^L\frac{s_l^y}{g_l^y}+{\displaystyle {\sum }_{m=1}^M\frac{s_m^b}{g_m^b})}}\right]}\right.\\ s.t.\left\{\begin{array}{l}{\displaystyle {\sum }_{t=1}^T{\displaystyle {\sum }_{c=1}^C{z}_c^t{x}_{cn}^t+s_n^x}}-x_{cn}^{t^{\prime }}=0,n=1,\dots ,N;\\ {\displaystyle {\sum }_{t=1}^T{\displaystyle {\sum }_{c=1}^C{z}_c^t{x}_{cl}^t-s_l^y}}-y_{cl}^{t^{\prime }}=0,l=1,\dots ,L;\\ {\displaystyle {\sum }_{t=1}^T{\displaystyle {\sum }_{c=1}^C{z}_c^t{x}_{cm}^t+s_m^b}}-b_{cm}^{t^{\prime }}=0,m=1,\dots ,M;\\ {\displaystyle {\sum }_{t=1}^T{\displaystyle {\sum }_{c=1}^C{z}_c^t}}=1,z_c^{t^{\prime }}\ge 0,c=1,\dots ,C;\\ s_n^x\ge 0,n=1,\dots ,N;s_l^y\ge 0,l=1,\dots ,L;s_m^b\ge 0,m=1,\dots ,M;\end{array}\right.\end{array}$$

(2)

where $\left(x_c^{t^{\prime }},y_c^{t^{\prime }},b_c^{t^{\prime }}\right)$ $x_c^{t^{\prime }}$, $y_c^{t^{\prime }}$ and $b_c^{t^{\prime }}$ denote the input and output vectors of the prefecture-level city $c$; $\left(g^x,g^y,g^b\right)$ denotes the direction vector; and $\left(s_n^x,s_i^y,s_m^b\right)$ denotes the slack vector.

(2)

GML index: The GML (Global Malmquist–Luenberger) index is based on the SBM directional distance function to measure the growth rate of GTFP based on the research idea of Oh et al. [29]. The GML index has circular additivity, which can reflect the long-term trend of GTFP while comparing the change in GTFP in each decision unit in the interim period, and its expression is as follows.

$$GM{L}_t^{t+1}=\frac{1+S^{*}\left(x^t,y^t,b^t;g\right)}{1+S^{*}\left(x^{t+1},y^{t+1},b^{t+1};g\right)},$$

(3)

The GML index reflects the change in GTFP in the current period compared to the previous period. An index greater than 1 indicates an increase in GTFP, while an index less than 1 indicates a decrease in GTFP.

3.2. Variable Selection and Data Description

3.2.1. Variable Selection

Explained variables: Drawing on relevant studies and the purpose of this paper, economic, energy and environmental [30] factors are taken into account to select reasonable input–output indicators. Furthermore, a system of urban GTFP measurement indicators was constructed (Table 1) and was set up as follows. (1) Input indicators were determined from three aspects: labor, capital and energy. They were characterized separately by the number of people employed in the whole society at the end of the year in each city (10,000 people), the stock of fixed assets (CNY 10,000,000) and the total annual electricity consumption (10,000,000 kilowatt hours). The perpetual inventory method is generally used to estimate the stock of fixed assets, and a lot of studies have been conducted by domestic and foreign scholars on this, such as Chow [31] and Chen et al. [16]. The basic formula is $K_{it}=K_{it-1}(1-{\delta }_{it})+I_{it}$. where $K$, $I$, and $\delta$ represent the capital stock, investment and depreciation rates, respectively. The depreciation rate was borrowed from the value of 10.96% measured by Shan [32]. The accurate measurement of labor factor input requires a comprehensive consideration of the quantity and quality of labor factors, which are often difficult to measure accurately. The idea of existing studies of primarily using the year-end employment number characterization was utilized. (2) Output indicators: The expected output was measured by the GDP of each region (billion USD) converted to 2013-based GDP to exclude the effect of inflation. Based on the primary control objects highlighted in the “13th Five-Year Plan” and related practices such as Liu [33,34,35], the non-expected outputs were characterized by industrial wastewater emissions, industrial waste gas emissions, carbon emissions, and PM2.5. In the overall inspector feedback, cities requiring air pollution rectification were selected as the experimental group. The most direct impact of the proposal to rectify air pollution is on the PM2.5 concentration in each city. As a result, the average PM2.5 value of each municipality is one of the non-expected outputs.

Control variables: In addition, considering the high-quality development of urban economy, other variables may be affected. In reference to the literature of the past few years [36,37], the following variables were added to the measurement model: (1) The population size (Popu) expressed as the city’s registered residence population at the end of the year. (2) The level of economic development (RGDP) measured by urban per capita GDP. (3) Financial development (Fin), for which we selected the ratio of the year-end deposit and loan balance of urban financial institutions to GDP. (4) The industrial structure (Indu) measured by the proportion of the secondary industry in terms of regional GDP. (5) Foreign direct investment (FDI), for which we used the amount of foreign investment actually utilized by the city.

Table 1. GTFP measurement index system.

Categories	Variables	Metrics
Input Indicators	Labor input Capital inputs Energy input	Year-end employment Fixed-asset investment (citywide) Total annual electricity consumption
Output Indicators	Desired output	GDP
	Non-desired output	Industrial wastewater emissions Industrial waste gas emissions Carbon emissions PM2.5

Table 1. GTFP measurement index system.

Categories	Variables	Metrics
Input Indicators	Labor input Capital inputs Energy input	Year-end employment Fixed-asset investment (citywide) Total annual electricity consumption
Output Indicators	Desired output	GDP
	Non-desired output	Industrial wastewater emissions Industrial waste gas emissions Carbon emissions PM2.5

3.2.2. Data Description

This paper uses the DID method, with GTFP as the explanatory variable measured only until 2019. In order to make the difference before and after the policy implementation more obvious, 2017 was selected as the time node of the central environmental protection inspection policy treatment. We took the first and second batch of inspected cities during and before 2017 as the experimental group. As there are new inspected cities every year after 2017, it would not be reasonable to use all non-inspected cities as the control group, so the new inspected cities during 2018–2019 from the control group were also excluded. In addition, the third batch of inspected cities with atmospheric problems announced their rectification plans in December 2017, and the actual effect of central environmental protection inspection was considered in 2018. Hence, they were not used for the experimental group. The 28 inspected cities are listed in Table 2.

This paper selects panel data from 252 prefecture-level cities in 30 Chinese provinces from 2013 to 2019 for empirical analysis (data from Hong Kong, Macao, Taiwan, and Tibet are excluded due to a serious lack of data). The data used are mainly from the China Regional Economic Statistical Yearbook and the China City Statistical Yearbook. In addition, the missing data for individual years are completed by interpolation.

Table 2. Selection of the central environmental protection inspection sample.

Province	Supervised Cities (First Batch)	Province	Supervised Cities (Second Batch)
Inner Mongolia	Huhehaote, Wuhai, Baotou, Chifeng	Hubei	Wuhan, Shiyan
Heilongjiang	Harbin, Heihe, Suihua, Mudanjiang	Shanxi	Xi’an, Xianyang
Jiangsu	Xuzhou, Lianyungang
Jiangxi	Nanchang, Xinyu, Yichun, Jiujiang, Ji’an, Pingxiang
Henan	Zhengzhou, Xinxiang, Kaifeng
Ningxia	Yinchuan, Shizuishan, Guyuan

Table 2. Selection of the central environmental protection inspection sample.

Province	Supervised Cities (First Batch)	Province	Supervised Cities (Second Batch)
Inner Mongolia	Huhehaote, Wuhai, Baotou, Chifeng	Hubei	Wuhan, Shiyan
Heilongjiang	Harbin, Heihe, Suihua, Mudanjiang	Shanxi	Xi’an, Xianyang
Jiangsu	Xuzhou, Lianyungang
Jiangxi	Nanchang, Xinyu, Yichun, Jiujiang, Ji’an, Pingxiang
Henan	Zhengzhou, Xinxiang, Kaifeng
Ningxia	Yinchuan, Shizuishan, Guyuan

3.3. Measurement Results

In order to further analyze the evolutionary characteristics of the GTFP of supervised cities and unsupervised cities, the trends of the GTFP of the two groups of cities from 2013 to 2019 was also simultaneously examined (see Figure 2), and superficial judgments were drawn as follows: (1) Except for 2017, during the period 2013–2019, the GTFP of cities nationwide showed an overall growth trend, and from 2013 to 2015, the 12th Five-Year Plan saw its final closing, and its main line of accelerating economic development achieved initial results. The transformation of the economic growth mode and the upgrading of the industrial structure led to an increase in the technological content of economic growth and promoted high-quality economic development. (2) The decline in green total factor productivity in 2017 was mainly due to the unprecedented efforts of the central environmental protection inspectors, who completed full coverage of 31 provinces in the two years between 2016 and 2017, with over 17,000 people held accountable. However, strict environmental remediation is not without cost; in the first round of environmental protection inspection alone, more than 200,000 chemical enterprises were legally shut down, and nearly 1,000,000 chemical enterprises underwent rectification. It is acknowledged that we have made significant improvements in air pollution prevention and control but also have a substantial negative impact on the economy in the short term. Considering that GDP is the core indicator of high-quality economic development, the strict enforcement has inevitably caused a significant drop in the city’s green total factor productivity.

4. Empirical Results and Analysis

According to the above analysis, since the selection of inspected cities is not random, but the result of selection, there is the problem of self-selection bias, and direct regression on the total sample may lead to errors in the results. Considering this reason, the PSM method is used to select the most similar cities to the experimental cities in the control group. Additionally, 28 inspected cities from the first and second batches of CEPI were selected as the experimental group, and the control group comprised non-inspected cities in 2017 and before, where inspections on cities between 2018 and 2019 were deducted. Population size (Popu), economic development level (RGDP), financial development (Fin), industrial structure (Indu) and foreign direct investment (FDI) were selected as matching variables, and propensity scores of the above matching indicators were estimated by logit models.

Figure 3 shows the density distribution of propensity scores for the experimental and control groups before and after matching. We can see that the distribution of propensity scores between the control and experimental groups after matching remained the same, indicating that the initial characteristics of the two groups of samples were similar in 2017. At the same time, according to the results of the matched balance test, it can be seen that the bias of the covariates of the two groups was effectively reduced, and all the matched variables were not significantly different. Therefore, the matched control group qualified as a counterfactual individual for the experimental group, and the PSM–DID method was feasible.

4.1. Analysis of the Impact of Central Environmental Protection Inspection on High-Quality Economic Development

After processing the research object, model (1) was used to empirically investigate the impact of central environmental protection inspection on GTFP. The specific estimation results of model (1) are shown in Table 3, where columns (1-1) are the regression results of the impact of central environmental protection inspection on GTFP before propensity score matching, and columns (1-2) are the regression results after PSM.

As can be seen from Table 3, after matching model (1), the coefficient of “time×treat” passed the significance level test above 5% and was positive. This shows that CEPI significantly impacted green total factor productivity and contributed to the high-quality development of the city’s economy. Public and media oversight of enterprises is strengthened to the maximum, thanks to the regular mechanism of public participation in environmental regulation established by the CEPI policy. As a broad social group, the public can keenly grasp the changes in the surrounding environment of the community, indirectly urge enterprises to invest in environmental protection, increase innovative research and development, improve the level of green innovation, and promote high-quality economic development [38]. (2) The relatively small impact of central environmental inspection on green total factor productivity suggests that the stringent “environmental storm” was not without cost. The closure and refurbishment of many chemical plants also had a short-lived impact on economic development, which was more evident in the cities with air pollution problems in the program. With GDP as the core indicator of high-quality economic development, the strict enforcement of environmental inspection has inevitably resulted in a modest increase in GTFP.

Table 3. Impact of central environmental protection inspection on high-quality economic development.

Variables	(1-1)	(1-2)
	DID	PSM–DID
CEPI × treat	0.007	0.013 **
	(0.005)	(0.007)
Popu	−0.012	−0.011
	(0.012)	(0.035)
RGDP	−0.02	0.05
	(0.06)	(0.11)
Indu	0.009	0.052 ***
	(0.008)	(0.013)
FDI	0.01	0.01 **
	(0.01)	(0.02)
Fin	−0.032	−0.021
	(0.032)	(0.027)
cons	1.139 ***	1.084 ***
	(0.098)	(0.210)
Time effect	Control	Control
Individual effect	Control	Control
N	1764	1412

Note: Values in brackets are standard errors: ** and *** indicate 5% and 1% significant levels, respectively.

Table 3. Impact of central environmental protection inspection on high-quality economic development.

Variables	(1-1)	(1-2)
	DID	PSM–DID
CEPI × treat	0.007	0.013 **
	(0.005)	(0.007)
Popu	−0.012	−0.011
	(0.012)	(0.035)
RGDP	−0.02	0.05
	(0.06)	(0.11)
Indu	0.009	0.052 ***
	(0.008)	(0.013)
FDI	0.01	0.01 **
	(0.01)	(0.02)
Fin	−0.032	−0.021
	(0.032)	(0.027)
cons	1.139 ***	1.084 ***
	(0.098)	(0.210)
Time effect	Control	Control
Individual effect	Control	Control
N	1764	1412

Note: Values in brackets are standard errors: ** and *** indicate 5% and 1% significant levels, respectively.

As for the control variables, the sign of each control variable did not change before and after matching, and the significance level of industrial structure and foreign direct investment increased after matching and passed the 1% and 5% significance tests, respectively. This suggests that industrial structure and foreign direct investment promote the improvement of GTF. Furthermore, the upgrading of industrial structure is often accompanied by the continued release of the “industrial structure dividend,” with the continued replacement of old and new industrial structures leading to the rise of efficient and clean industries. The inflow of FDI provides local enterprises with an incentive to innovate. In the face of advanced foreign companies of the same type, Chinese enterprises are motivated to learn advanced technologies and invest more in R&D and innovation. A win–win situation of enhancing competitiveness and technological progress is achieved, promoting high-quality economic development.

4.2. Analysis of Dynamic Effects

The result of DID satisfies consistency provided that the experimental and control groups satisfy the parallel trend hypothesis, i.e., there is no significant difference in the trend of the outcome variable in the experimental group in the absence of policy interventions that are largely consistent. Accordingly, this paper follows Ren Shenggang et al. [39] in treating the grouped dummy variable with a cross-product term for all sample years. In addition, the above regression results reflect the average impact of the pilot policy implementation on urban green total factor productivity. They do not reflect the differences in this impact across time for the pilot policy. For this reason, the event study approach empirically tests the pilot policy’s dynamic effects and constructs the following model.

$$Y_{it}={\delta }_0+{\displaystyle \sum _{n=2013}^{2019}{\beta }_n}CEP{I}_{it}\times yea{r}_{it}+{\delta }_1{X}_{it}+{\mu }_i+{\lambda }_i+{\epsilon }_{it},$$

(4)

In order to avoid the problem of complete collinearity, period 1 before the policy implementation point was chosen as the baseline group, and thus data for period −1 are not available in the figure, denoting a range of estimates from 2013 to 2019. Other variables are defined in the regression model Equation (3).

Figure 4 plots the estimation results of ${\beta }_n$ at a 95% confidence interval, which shows that ${\beta }_n$ was not significant from 2013 to 2016, indicating no significant difference between inspected cities and non-inspected cities before the experiment, satisfying the parallel trend hypothesis. In addition, ${\beta }_n$ only had a significant positive impact in 2017, the year of policy implementation, indicating that CEPI has achieved immediate results through its strict enforcement, and that environmental pollution has been dramatically improved, especially in the aspect of air pollution. However, the improvement results will gradually weaken when the inspection team leaves. This is mainly because the current economy has entered the new normal, the effects of various macro policies are getting worse and worse, the foreign trade situation is complex and changeable, and a large number of enterprises have stopped working for rectification. In order to avoid being punished under high-intensity environmental pressure, some local governments may even adopt “all shutdown “or “shutdown instead of governance”, which directly slows down economic growth and further affects the significant improvement of GTFP [40]. In addition, some persistent problems of perfunctory rectification, superficial rectification, and incomplete rectification still exist. CEPI should both strengthen the constraint mechanism and innovate the incentive mechanism, and the next step should be to improve on how to gain long-term policy influence.

4.3. Regional Heterogeneity Test

Table 4. Impact of central environmental protection inspection on the high-quality economic development of cities in different regions.

	(1)	(2)	(3)
Variables	Eastern	Central	Western
CEPI × treat	0.042 **	0.013	−0.023 **
	(0.027)	(0.012)	(0.015)
R2	0.749	0.613	0.539
Time effect	Control	Control	Control
Individual effect	Control	Control	Control
N	693	553	166

Note: Values in brackets are standard errors: ** indicate 5% significant levels.

shows that, after controlling for time and individual fixed effects, central environmental inspection significantly has contributed to high-quality economic development in the eastern region, but has had a significant dampening effect in the western region. The reason is that eastern cities have a higher level of marketization and a better and more systematic innovation environment that is more conducive to the implementation of environmental policies, as well as to attracting further foreign capital inflows and promoting an advanced industrial structure that contributes to high-quality economic development [41]. For the central cities, the central environmental policy did not have a significant impact on the central cities, mainly due to their abundant energy endowment and national industrial policies, which have taken over most of the labor-intensive and resource-intensive industries, and the total energy consumption and pollution emissions have increased in parallel with economic growth. For the western region, as enterprises in the west are inherently less innovative, when they encounter CEPI, the increased costs of enterprises cannot compensate for the benefits of innovation, causing their production and operation activities to suffer. In addition, the high reliance on resource and energy development makes it difficult for the green effect to compensate for the loss of economic development, so the high quality of economic development in western cities is affected to some extent by the central environmental protection inspection policy.

4.4. The Placebo Test

A placebo test was conducted by randomly assigning pilot cities to test whether unobservable factors drove these results. Specifically, 28 cities were randomly selected as the treatment group from 252 prefecture-level cities, which were assumed to be cities with air pollution problems. The rest of the region was the control group. Random sampling ensures that the independent variable $CEP{I}_{}\times yea{r}_{}$ constructed in this paper does not affect urban green total factor productivity. In other words, any significant finding would indicate a bias in the regression results of this paper. Additionally, 1000 random samples were conducted, and the baseline regressions were run according to Equation (1). Figure 5 reports the means of the regression estimates after 1000 random assignments. It is further discovered that the mean of the estimated coefficients for all $CEP{I}_{}\times yea{r}_{}$ variables is almost zero, much smaller than the true estimate of 0.013, marked with a red line in the figure, suggesting that the estimates are unlikely to be driven by unobservable factors in cities and years.

5. Key Conclusions and Implications

Based on the above analysis, the conclusions are drawn as follows:

(1) Overall, CEPI has promoted high-quality economic development through technological innovation and resource allocation, and parallel trend tests and placebo tests support this result.

(2) On a regional basis, CEPI significantly promotes high-quality economic development in eastern cities but has a significant dampening effect in western regions, mainly due to differences in regional industrial structure levels, resource endowment conditions, and enterprise innovation levels.

(3) Analysis of the dynamic effects of the policy suggests that CEPI did not have a sustained impact on high-quality economic development, and problems such as data falsification and ineffective rectification still exist.

Based on the above research, the implications of this paper are as follows:

(1) There is no contradiction between strengthening environmental regulations and improving the quality of economic development. The key lies in using flexible and appropriate environmental regulations to give enterprises continuous incentives for innovation. In particular, macro policies are becoming less effective as the economy enters a new standard. The foreign trade situation is complex and volatile. Appropriate incentives for innovation, such as the establishment of assessment mechanisms and reward and punishment systems, can not only promote the transformation of polluting enterprises and high-quality economic development, but also reduce the cost of government supervision. For this reason, the central government needs to focus on how to do a good job of top-level design at the national level and encourage bold local practices and breakthroughs, using a variety of tools to patiently build and polish a “five-in-one” pattern of the harmonious development of the economy, environment and society.

(2) Strengthening environmental enforcement ensures the effective implementation of the CEPI system. This paper finds that the central environmental protection inspection system only has a significant impact in the year of policy implementation and lacks a sustained and practical impact in later years. In this regard, the Chinese government should further strengthen environmental enforcement and establish a sound monitoring and tracking mechanism to eliminate possible perfunctory, superficial, and pretend rectification.

(3) Set inspection standards of different intensities according to local conditions. The eastern region has a better industrial structure and more muscular economic strength. CEPI can help industries eliminate backward production capacity and achieve an advanced industrial structure. Therefore, it is possible to appropriately increase the intensity of inspections, strengthen enforcement and supervision, and improve legal provisions related to environmental protection. Furthermore, the intensity of inspections in the western region, where large-scale rectification is bound to impact economic development, should be moderately reduced.

(4) As a policy with environmental governance characteristics in China, CEPI is explored and practiced in response to the needs of the people in China. Therefore, the theme of this article is very hermetic, and it is difficult to find the same policy in other countries and obtain a more comprehensive result. In the future, we will pay more attention to other countries’ environmental conditions and policies and draw links with them to learn from them.